tag:old.nabble.com,2006:forum-730Nabble - Berkeley DB Xml2007-06-20T00:31:06ZBerkeley DB XML is the native XML database engine for your product. Berkeley DB XML provides XQuery access into a database of document containers. Berkeley DB Xml home is <a href="http://www.sleepycat.com/" target="_top" rel="nofollow">here</a>.tag:old.nabble.com,2006:post-11207897BDB daemon2007-06-20T00:31:06Z2007-06-20T00:31:06Zsrini19hi,
I am new to BDB and was wondering how to run a BDB daemon process.
thanks,
srini.<p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-10239346Re: [bdbxml] XPATH_EVALUATION_ERROR2007-04-29T00:20:33Z2007-04-29T00:20:33Zahlandberg<br><blockquote class="quote light-black dark-border-color"><div class="quote light-border-color">
<div class="quote-author" style="font-weight: bold;">George Feinberg wrote:</div>
<div class="quote-message shrinkable-quote">Maike,
<br>> I have a problem with query XML-documents. I can create and open
<br>> containern, put documents into the container and get documents from
<br>> the container. But when I use the query
<br>>
<br>> collection('D:/myContainer.bdbxml')/*
<br>>
<br>> I get the following error:
<br>>
<br>> com.sleepycat.dbxml.XmlException: Error: Cannot resolve collection:
<br>> D:/myContainer.bdbxml, errcode = XPATH_EVALUATION_ERROR
<br><br>This is a known issue in release 2.1.8. Using absolute paths
<br>as an argument to the collection() function on Windows
<br>fails to resolve the argument to a BDB XML container.
<br><br>The simplest workaround is to alias the container. After you
<br>open the container, use this method:
<br> container.addAlias("mycontainer");
<br>Then, in the query, you can use:
<br> collection('mycontainer')/*
<br><br>In general, aliasing container names is useful to eliminate
<br>all possibility of different paths resolving to different container
<br>names. It is no less efficient, internally, to use aliases.
<br><br>Regards,
<br><br>George
<br></div>
</div></blockquote>
<br>Hello.
<br><br>I had exactly the same problem. However, the workaround didn't work for me. I have BDBXML 2.3.10.
<br><br>I have tried the following:
<br><br>I created an alias for my container
<br><br>myContainer.addAlias("myContainerAlias");
<br><br>Then I set the base uri to empty, after realising that giving it the absolute path was useless:
<br><br>queryContext.setBaseURI( containerURL );
<br><br>and also
<br><br>queryContext.setBaseURI( "" );
<br><br>Then I set my query to:
<br><br>String myQuery = "collection('myContainerAlias')/phonebook";
<br><br>But I got this error:
<br><br>Query Expression: collection('myContainerAlias')/phonebook
<br>com.sleepycat.dbxml.XmlException: Error: External access not allowed. Cannot resolve
<br> collection: D:/dev/workspace/BDBXML/xml_containers/myContainerAlias, errcode = XPATH_EVALUATION_ERROR
<br><br><br>What's wrong here?
<br><br>Cheers, Anders<p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-8941292Maven2 support2007-02-13T01:50:58Z2007-02-13T01:50:58ZshirazHi
<br>Is it possible to use Maven2 with BerkeleyDBXML while developing in java?
<br>Thanks for any help in advance.
<br>regards
<br>shiraz<p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-8182608deleteDocument doesn't work2007-01-05T09:51:34Z2007-01-05T09:51:34Zgiancarlohallo,
<br>I have a problem. I have created 2 Java class for save and delete XMLDocument, but the delete Class is not working:
<br><br>class DeleteDocumentCommand extends AbstractCommand {
<br><br> private String name;
<br> private XmlUpdateContext updateContext;
<br> private XmlTransaction txn;
<br><br> public DeleteDocumentCommand(DbEnv dbEnv,String name) {
<br> super(dbEnv);
<br> this.name = name;
<br> }
<br><br> public void close() throws CommandException {
<br> try {
<br> if(txn!=null){
<br> txn.commit();
<br> txn.delete();
<br> }
<br> if(updateContext!=null){
<br> updateContext.delete();
<br> }
<br> } catch (XmlException e) {
<br> e.printStackTrace();
<br> throw new CloseCommandException(e);
<br> }
<br> }
<br><br> public Object execute() throws CommandException {
<br> try{
<br> openedContainer.deleteDocument(txn,name, updateContext);
<br> return null;
<br> }catch(Exception e){
<br> e.printStackTrace();
<br> throw new ExecuteCommandException(e);
<br> }
<br> }
<br><br> public void open() throws CommandException {
<br> theMgr = dbEnv.getManager();
<br> XmlContainerConfig config = new XmlContainerConfig();
<br> config.setTransactional(true);
<br> try {
<br> openedContainer = theMgr.openContainer(XMLDBConstant.theContainer, config);
<br> updateContext = theMgr.createUpdateContext();
<br> txn = theMgr.createTransaction();
<br> } catch (XmlException e) {
<br> throw new OpenCommandException(e);
<br> }
<br> }
<br><br>}
<br><br><br>class SaveDocumentCommand extends AbstractCommand {
<br>
<br> private String name;
<br> private String content;
<br> private XmlUpdateContext updateContext;
<br> private XmlTransaction txn;
<br><br> public SaveDocumentCommand(DbEnv dbEnv,String name,String content) {
<br> super(dbEnv);
<br> this.name = name;
<br> this.content = content;
<br> // TODO Auto-generated constructor stub
<br> }
<br><br> public void close() throws CommandException {
<br> try {
<br> if(txn!=null){
<br> txn.commit();
<br> txn.delete();
<br> }
<br> if(updateContext!=null){
<br> updateContext.delete();
<br> }
<br> } catch (XmlException e) {
<br> throw new CloseCommandException(e);
<br> }
<br> }
<br><br> public Object execute() throws CommandException {
<br> try{
<br> XmlDocument xmlDoc = theMgr.createDocument();
<br> xmlDoc.setContent(content);
<br> xmlDoc.setName(name);
<br> Date theDate = new Date();
<br> xmlDoc.setMetaData(XMLDBConstant.uri, XMLDBConstant.name, new XmlValue(theDate.toString()));
<br> openedContainer.putDocument(txn, xmlDoc, updateContext);
<br> return null;
<br> }catch(Exception e){
<br> throw new ExecuteCommandException(e);
<br> }
<br>
<br> }
<br><br> public void open() throws CommandException {
<br> theMgr = dbEnv.getManager();
<br> XmlContainerConfig config = new XmlContainerConfig();
<br> config.setTransactional(true);
<br> try {
<br> openedContainer = theMgr.openContainer(XMLDBConstant.theContainer, config);
<br> updateContext = theMgr.createUpdateContext();
<br> txn = theMgr.createTransaction();
<br> } catch (XmlException e) {
<br> throw new OpenCommandException(e);
<br> }
<br>
<br>
<br> }
<br><br><br>public class DbEnv
<br>{
<br>
<br><br> private Environment dbEnv_ = null;
<br> private XmlManager mgr_ = null;
<br> private File path2DbEnv_ = null;
<br>
<br> public static boolean exist = false;
<br><br> public DbEnv() throws Exception {
<br> File path2DbEnv = new File(XMLDBConstant.path2DbEnvString);
<br> if (! path2DbEnv.isDirectory()) {
<br> throw new Exception(path2DbEnv.getPath() +
<br> " does not exist or is not a directory.");
<br> }
<br><br> EnvironmentConfig config = new EnvironmentConfig();
<br> config.setCacheSize(50 * 1024 * 1024); // 50MB
<br> config.setAllowCreate(true);
<br> config.setInitializeCache(true);
<br> config.setTransactional(true);
<br> config.setInitializeLocking(true);
<br> config.setInitializeLogging(true);
<br> config.setErrorStream(System.err);
<br> dbEnv_ = new Environment(path2DbEnv, config);
<br><br> path2DbEnv_ = path2DbEnv;
<br> mgr_ = new XmlManager(dbEnv_, null);
<br> if(!exist){
<br> if(mgr_.existsContainer(XMLDBConstant.theContainer)==0){
<br> System.out.println("Sto creando il DB");
<br> mgr_.createContainer(XMLDBConstant.theContainer);
<br> }
<br> exist=true;
<br> }
<br> }
<br>
<br>
<br> public File getDbEnvPath() { return path2DbEnv_; }
<br><br> public Environment getEnvironment() { return dbEnv_; }
<br><br> public XmlManager getManager() { return mgr_; }
<br><br> public void cleanup() throws DatabaseException
<br> {
<br> if (mgr_ != null) {
<br> mgr_.delete();
<br> mgr_ = null;
<br> }
<br> if (dbEnv_ != null) {
<br> dbEnv_.close();
<br> dbEnv_ = null;
<br> }
<br><br> }
<br><br><br>I'm using my class:
<br><br>Command command = new DeleteDocumentCommand (new DbEnv(),"name.xml");
<br>command.open();
<br>command.execute();
<br>command.close();
<br><br>The method openedContainer.deleteDocument(txn,name, updateContext) seems blocked. I waited for, but it does not go ahead;
<br><br>Can you help me?
<br>Thanks
<br>giancarlo<p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-7126899setname or value(string)error2006-11-01T18:49:52Z2006-11-01T18:49:52Zlee yeonbokhello,
<br>I hope you will allow my bad English.
<br>I am trying to make a tool that retrieve XML files by Xquery with BDBXML in Visual Studio c++.net 2003.
<br>I always get this error message when I run the tool.
<br><br>「Debug Assertion failed!
<br><br>Program: .........
<br>File:dbgdel.cpp
<br>Line:52
<br><br>Expression:_BLOCK_TYPE_IS_VALID(pHead->nBlockUse)
<br><br>For information on how your program can cause an assertion failure, see the Visual C++ documentation on asserts.」
<br><br><br>This error message appears in these cases:
<br>1. using putDocument with document name consist of upper 16 letters.
<br>2. the return value (string) is consist of upper 16 letters.
<br><br>In both cases, the tool runs well if the data are consist of 15 letters or below.
<br><br>There may be more cases that shows the error message.
<br>I guess there is a trouble in dealing with string data.
<br>However I cannot find strange codes in APIs.
<br><br>Is anybody who has similar problem with me?
<br>I would be very pleased if I get helps.
<br><br><br>regards,
<br>Lee
<br> <p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5761345RE: Unexpected Tomcat closure breaks containers?2006-08-11T06:18:42Z2006-08-11T06:18:42ZAntony GrinyerGeorge,
<br><br>I'll move this issue to the new forum - thanks.
<br><br>Ant
<br><div class='shrinkable-quote'><br>> -----Original Message-----
<br>> From: George Feinberg [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5761345&i=0" target="_top" rel="nofollow">george.feinberg@...</a>]
<br>> Sent: 10 August 2006 19:42
<br>> To: Antony Grinyer
<br>> Cc: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5761345&i=1" target="_top" rel="nofollow">xml@...</a>; John Snelson
<br>> Subject: Re: [bdbxml] Unexpected Tomcat closure breaks containers?
<br>>
<br>>
<br>> Antony,
<br>>
<br>> What configuration do you use to create the container itself?
<br>>
<br>> Reminder -- this list could start bouncing email any day now.
<br>> Traffic is moving to the OTN forum:
<br>> <a href="http://forums.oracle.com/forums/forum.jspa?forumID=274" target="_top" rel="nofollow">http://forums.oracle.com/forums/forum.jspa?forumID=274</a><br>>
<br>> George
<br>>
<br>> > Having followed all the advice to recover the state of our
<br>> containers
<br>> > our container still won't open after a Tomcat crash despite
<br>> using the
<br>> > run recovery parameters; therefore I guess as a last
<br>> effort, it's best
<br>> > to describe the setup and scenario we have.
<br>> >
<br>> > We use DBXML 2.2.13 and set the following environment
<br>> parameters when
<br>> > our application starts:
<br>> >
<br>> > EnvironmentConfig config = new EnvironmentConfig();
<br>> > config.setAllowCreate(true); config.setInitializeCache(true);
<br>> > config.setInitializeLocking(true);
<br>> > config.setInitializeLogging(true);
<br>> > config.setTransactional(true);
<br>> > config.setRunRecovery(true);
<br>> > config.setVerboseRecovery(true);
<br>> > config.setCacheSize(512000000);
<br>> > config.setMaxLocks(2000);
<br>> > config.setMaxLockers(2000);
<br>> > config.setMaxLockObjects(2000);
<br>> >
<br>> > We use Tomcat 5.5.17 as our application server. To simulate the
<br>> > problem I run Tomcat which in turn starts our application;
<br>> I login to
<br>> > our application and create a container which stays open for the
<br>> > lifetime of the application (while Tomcat is running); then, if we
<br>> > close the application altogether and close Tomcat, we have
<br>> a listener
<br>> > hooked into Tomcat which closes all the open containers and the
<br>> > XmlManager we use in the application. This seems to work
<br>> very nicely
<br>> > if we close Tomcat sensibly. After simulating a Tomcat
<br>> crash (killing
<br>> > the java.exe process in Task Manager, which does not close the
<br>> > containers or XmlManager object gracefully) we log into our
<br>> > application and try to select the container we created -
<br>> this results
<br>> > in an error message reported to the console saying the
<br>> container does
<br>> > not exist; interestingly, we also get a reported message that a
<br>> > recovery was complete. Here are the messages written to the console:
<br>> >
<br>> > Finding last valid log LSN: file: 1 offset 16856 Recovery
<br>> complete at
<br>> > Wed Aug 09 11:06:56 2006 Maximum transaction ID 0 Recovery
<br>> checkpoint
<br>> > [0][0]
<br>> > com.sleepycat.dbxml.XmlException: Error: No such file or directory,
<br>> > errcode = DATABASE_ERROR
<br>> >
<br>> > The 'No such file or directory' error occurs even though the
<br>> > containers exist.
<br>> >
<br>> > Note we are not using transactions in our code; however
<br>> we're not sure
<br>> > how this would prevent the same problem as the container
<br>> would still
<br>> > be open during the crash and subsequently would become
<br>> > unstable/corrupted without a sensible closure.
<br>> >
<br>> > Because of this, we're thinking of implementing some sort
<br>> of real-time
<br>> > back up plan (in addition to our robust incremental backup plan) to
<br>> > copy the containers to another directly every so often to ensure we
<br>> > can at least pull back the containers quickly.
<br>> >
<br>> > Any last ditch advice would be great.
<br>> >
<br>> > Thanks,
<br>> > Ant
<br>> > CSW Group Ltd.
<br>>
<br>>
<br>>
<br>> ------------------------------------------
<br>> To remove yourself from this list, send an email to
<br>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5761345&i=2" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>
<br>>
<br>>
</div><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5761345&i=3" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5750197Re: Unexpected Tomcat closure breaks containers?2006-08-10T12:41:51Z2006-08-10T12:41:51ZGeorge Feinberg-3<br>Antony,
<br><br>What configuration do you use to create the container itself?
<br><br>Reminder -- this list could start bouncing email any day now.
<br>Traffic is moving to the OTN forum:
<br> <a href="http://forums.oracle.com/forums/forum.jspa?forumID=274" target="_top" rel="nofollow">http://forums.oracle.com/forums/forum.jspa?forumID=274</a><br><br>George
<br><div class='shrinkable-quote'><br>> Having followed all the advice to recover the state of our containers
<br>> our container still won't open after a Tomcat crash despite using the
<br>> run recovery parameters; therefore I guess as a last effort, it's best
<br>> to describe the setup and scenario we have.
<br>>
<br>> We use DBXML 2.2.13 and set the following environment parameters when
<br>> our application starts:
<br>>
<br>> EnvironmentConfig config = new EnvironmentConfig();
<br>> config.setAllowCreate(true);
<br>> config.setInitializeCache(true);
<br>> config.setInitializeLocking(true);
<br>> config.setInitializeLogging(true);
<br>> config.setTransactional(true);
<br>> config.setRunRecovery(true);
<br>> config.setVerboseRecovery(true);
<br>> config.setCacheSize(512000000);
<br>> config.setMaxLocks(2000);
<br>> config.setMaxLockers(2000);
<br>> config.setMaxLockObjects(2000);
<br>>
<br>> We use Tomcat 5.5.17 as our application server. To simulate the
<br>> problem
<br>> I run Tomcat which in turn starts our application; I login to our
<br>> application and create a container which stays open for the
<br>> lifetime of
<br>> the application (while Tomcat is running); then, if we close the
<br>> application altogether and close Tomcat, we have a listener hooked
<br>> into
<br>> Tomcat which closes all the open containers and the XmlManager we
<br>> use in
<br>> the application. This seems to work very nicely if we close Tomcat
<br>> sensibly. After simulating a Tomcat crash (killing the java.exe
<br>> process
<br>> in Task Manager, which does not close the containers or XmlManager
<br>> object gracefully) we log into our application and try to select the
<br>> container we created - this results in an error message reported to
<br>> the
<br>> console saying the container does not exist; interestingly, we also
<br>> get
<br>> a reported message that a recovery was complete. Here are the messages
<br>> written to the console:
<br>>
<br>> Finding last valid log LSN: file: 1 offset 16856
<br>> Recovery complete at Wed Aug 09 11:06:56 2006
<br>> Maximum transaction ID 0 Recovery checkpoint [0][0]
<br>> com.sleepycat.dbxml.XmlException: Error: No such file or directory,
<br>> errcode = DATABASE_ERROR
<br>>
<br>> The 'No such file or directory' error occurs even though the
<br>> containers
<br>> exist.
<br>>
<br>> Note we are not using transactions in our code; however we're not sure
<br>> how this would prevent the same problem as the container would
<br>> still be
<br>> open during the crash and subsequently would become unstable/corrupted
<br>> without a sensible closure.
<br>>
<br>> Because of this, we're thinking of implementing some sort of real-time
<br>> back up plan (in addition to our robust incremental backup plan) to
<br>> copy
<br>> the containers to another directly every so often to ensure we can at
<br>> least pull back the containers quickly.
<br>>
<br>> Any last ditch advice would be great.
<br>>
<br>> Thanks,
<br>> Ant
<br>> CSW Group Ltd.
</div><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5750197&i=0" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5723260RE: Unexpected Tomcat closure breaks containers?2006-08-09T04:20:27Z2006-08-09T04:20:27ZAntony GrinyerHaving followed all the advice to recover the state of our containers
<br>our container still won't open after a Tomcat crash despite using the
<br>run recovery parameters; therefore I guess as a last effort, it's best
<br>to describe the setup and scenario we have.
<br><br>We use DBXML 2.2.13 and set the following environment parameters when
<br>our application starts:
<br><br>EnvironmentConfig config = new EnvironmentConfig();
<br>config.setAllowCreate(true);
<br>config.setInitializeCache(true);
<br>config.setInitializeLocking(true);
<br>config.setInitializeLogging(true);
<br>config.setTransactional(true);
<br>config.setRunRecovery(true);
<br>config.setVerboseRecovery(true);
<br>config.setCacheSize(512000000);
<br>config.setMaxLocks(2000);
<br>config.setMaxLockers(2000);
<br>config.setMaxLockObjects(2000);
<br><br>We use Tomcat 5.5.17 as our application server. To simulate the problem
<br>I run Tomcat which in turn starts our application; I login to our
<br>application and create a container which stays open for the lifetime of
<br>the application (while Tomcat is running); then, if we close the
<br>application altogether and close Tomcat, we have a listener hooked into
<br>Tomcat which closes all the open containers and the XmlManager we use in
<br>the application. This seems to work very nicely if we close Tomcat
<br>sensibly. After simulating a Tomcat crash (killing the java.exe process
<br>in Task Manager, which does not close the containers or XmlManager
<br>object gracefully) we log into our application and try to select the
<br>container we created - this results in an error message reported to the
<br>console saying the container does not exist; interestingly, we also get
<br>a reported message that a recovery was complete. Here are the messages
<br>written to the console:
<br><br>Finding last valid log LSN: file: 1 offset 16856
<br>Recovery complete at Wed Aug 09 11:06:56 2006
<br>Maximum transaction ID 0 Recovery checkpoint [0][0]
<br>com.sleepycat.dbxml.XmlException: Error: No such file or directory,
<br>errcode = DATABASE_ERROR
<br><br>The 'No such file or directory' error occurs even though the containers
<br>exist.
<br><br>Note we are not using transactions in our code; however we're not sure
<br>how this would prevent the same problem as the container would still be
<br>open during the crash and subsequently would become unstable/corrupted
<br>without a sensible closure.
<br><br>Because of this, we're thinking of implementing some sort of real-time
<br>back up plan (in addition to our robust incremental backup plan) to copy
<br>the containers to another directly every so often to ensure we can at
<br>least pull back the containers quickly.
<br><br>Any last ditch advice would be great.
<br><br>Thanks,
<br>Ant
<br>CSW Group Ltd.
<br><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5723260&i=0" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5711062new book, quick reference, and alternate mailing list2006-08-08T11:14:37Z2006-08-08T11:14:37ZDanny BrianBefore the list is terminated, I wanted to let subscriber know that
<br>"The Definitive Guide to Berkeley DB XML" will be available from
<br>Apress later this month.
<br><br> <a href="http://apress.com/book/bookDisplay.html?bID=10128" target="_top" rel="nofollow">http://apress.com/book/bookDisplay.html?bID=10128</a><br><br>George Feinberg did a lot of work on this; thanks again George.
<br><br>In part to promote the book, I've made a quick API reference sheet
<br>for BDB XML available for download from:
<br><br> <a href="http://conceptuary.com/bdbxml/" target="_top" rel="nofollow">http://conceptuary.com/bdbxml/</a><br><br>And lastly, my employer is hosting an unofficial BDB XML mailing list
<br>(described at the above link) for users who want to continue
<br>discussions via email.
<br><br>Regards,
<br>Danny
<br><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5711062&i=0" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5708022Berkeley DB mailing list migration.2006-08-08T08:43:13Z2006-08-08T08:43:13ZKeith Bostic-2All of the Sleepycat Software Berkeley DB support channels will move
<br>to Oracle support channels over the next few days.
<br><br>This includes the Berkeley DB Java Edition, Berkeley DB XML and
<br>Berkeley DB High Availability product mailing lists.
<br><br>The mailing lists are being moved to the Oracle Technical Network
<br>forums:
<br><br> <a href="http://forums.oracle.com/forums/category.jspa?categoryID=18" target="_top" rel="nofollow">http://forums.oracle.com/forums/category.jspa?categoryID=18</a><br><br>There are four technical forums at this site: one each for Berkeley
<br>DB, Berkeley DB High Availability, Berkeley DB XML and Berkeley DB
<br>Java Edition. Please select the appropriate forum and post your
<br>question there.
<br><br>The Berkeley DB support and engineering teams are already monitoring
<br>these forums, and you can expect prompt and helpful answers.
<br><br>Regards,
<br>Keith Bostic
<br><br>=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
<br>Keith Bostic
<br>+1-781-259-3139
<br>keithbosticim (ymsgid)
<br><a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5708022&i=0" target="_top" rel="nofollow">keith.bostic@...</a>
<br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5708022&i=1" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5706918Poor performance after index add2006-08-08T07:45:19Z2006-08-08T07:45:19ZBreno CostaHello everybody,<br><br> I'm puzzled this time. I've got a container with about 1700 small xml documents, each of which have this structure:<br><br><?xml version="1.0" encoding="UTF-8"?><br><BO>
<br> <cadastro><br> <id>1673</id><br> <ano>06</ano><br> <hash>dd8897c6f8556cb455ab581954e51072</hash><br> <inicio><br> <data><br> <d>07</d><br> <m>08</m>
<br> <a>2006</a><br> <date>2006-08-07</date><br> </data><br> <hora><br> <h>10</h><br> <m>35</m><br> <s>27</s><br> <time>10:35:27</time>
<br> </hora><br> </inicio><br> <fim><br> <data><br> <d>07</d><br> <m>08</m><br> <a>2006</a><br> <date>2006-08-07</date><br> </data>
<br> <hora><br> <h>10</h><br> <m>58</m><br> <s>23</s><br> <time>10:58:23</time><br> </hora><br> </fim><br> <numFlash/><br> <remoteAddr>
<a href="http://200.254.130.125" target="_top" rel="nofollow">200.254.130.125</a></remoteAddr><br> <remotePort>1329</remotePort><br> <maquina/><br> <sispol/><br> </cadastro><br> <validacao><br> <situacao>PROCESSAMENTO</situacao>
<br> <validadores><br> <avaliador><br> <nome>José Lucivan Gomes de Omena</nome><br> <cpf><br> <numero>77609093491</numero><br> </cpf><br> <matricula/><br></avaliador><escrivao>
<br> <nome/><br> <matricula/><br> <cpf><br> <numero>00000000000</numero><br> </cpf><br> </escrivao><br> <delegado><br> <nome/><br> <matricula/>
<br> <cpf><br> <numero>00000000000</numero><br> </cpf><br> </delegado><br> </validadores><br> <data><br> <d>0</d><br> <m>0</m><br> <a>0</a>
<br> <date>0-0-0</date><br> </data><br> <hora><br> <h>00</h><br> <m>00</m><br> <s>00</s><br> <time>00:00:00</time><br> </hora><br>
<maquina><br> <dhCadastro><br> <data><br> <d>0</d><br> <m>0</m><br> <a>0</a><br> <date>0-0-0</date><br> </data><br> <hora>
<br> <h>00</h><br> <m>00</m><br> <s>00</s><br> <time>00:00:00</time><br> </hora><br> </dhCadastro><br> <nome/><br> <descricao/>
<br> <ip/><br> </maquina><br> <obs/><br> </validacao><br> <fato><br> <natureza>EXTRAVIO</natureza><br> <uf>AL</uf><br> <cidade>MACEIÓ</cidade><br> <bairro>SANTA AMÉLIA</bairro>
<br> <lougradouro/><br> <pontoReferencia/><br> <data><br> <d>01</d><br> <m>08</m><br> <a>2006</a><br> <date>2006-08-01</date><br> </data>
<br> <periodo>tarde</periodo><br> <descricaoFato/><br> <outrosDocumentos/><br> </fato><br> <pessoas><br> <vitima><br> <cpf><br> <numero>01079983430</numero>
<br> </cpf><br> <rg><br> <orgaoEmissor>SSP</orgaoEmissor><br> <uf>AL</uf><br> <dExpedicao><br> <d>0</d><br> <m>0</m><br> <a>0</a>
<br> <date>0-0-0</date><br> </dExpedicao><br> <numero>1933214</numero><br> </rg><br> <nome>JAMILLE SAMPAIO CALHEIROS</nome><br> <sexo>F</sexo><br>
<dNasc><br> <d>12</d><br> <m>09</m><br> <a>1982</a><br> <date>1982-09-12</date><br> </dNasc><br> <lougradouro>RUA HILDA DE MENEZES BARROS</lougradouro>
<br> <numero>71</numero><br> <bairro>SANTA AMÉLIA</bairro><br> <compl>CJ COLINAS</compl><br> <cidade>MACEIÓ</cidade><br> <uf>AL</uf><br> <cep>57063450</cep>
<br> <naturalidade><br> <cidade>MACEIÓ</cidade><br> <uf>AL</uf><br> </naturalidade><br> <filiacao><br> <pai>CICERO DA ROCHA CALHEIROS</pai><br> <mae>MARCIA MARIA SAMPAIO CALHEIROS</mae>
<br> </filiacao><br> <fone><br> <numResidencial>8232238810</numResidencial><br> <numCelular/><br> </fone><br> <email/><br> <estadoCivil>1</estadoCivil>
<br> <grauInstrucao>5</grauInstrucao><br> <turista>n</turista><br> </vitima><br> </pessoas><br> <documentos><br> <ctps><br> <numero>047758</numero><br> <serie>00019</serie>
<br> </ctps><br> </documentos><br> <objetos/><br> <historico>O DECLARANTE INFORMA QUE, NO LOCAL, DATA E HORA ACIMA CITADOS, FORA EXTRAVIADO O SEGUINTE DOCUMENTO: CTPS, NÚMERO 047758, SÉRIE 00019; OBS.: ESTE BOLETIM NÃO SUBSTITUI O(S) DOCUMENTO(S) ACIMA CITADO(S).</historico>
<br></BO><br><br><br> This container has no index at all, and the performance is, up to now, very good. Of course this might change as the container gets bigger, so that's why I'm adding indexes now. I just did the following, from the shell:
<br><br>openContainer BOs.dbxml<br>addIndex "" natureza node-element-equality-string<br>addIndex "" cpf node-element-equality-string<br>addIndex "" remoteAddr node-element-equality-string<br>
addIndex "" id node-element-equality-string<br>addIndex "" ano node-element-equality-string<br>quit<br><br> These are the most frequent searches. But after doing this, the system (which is a web app) got very, very slow, sometimes the JVM even crashes.
<br><br> So, what could be going wrong with this? I appreciate any help.<br><br>cheers,<br><br>-- Breno Jacinto<br><br>
<p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5633572RE: Unexpected Tomcat closure breaks containers?2006-08-03T08:16:58Z2006-08-03T08:16:58ZAntony GrinyerGeorge - yes we have EnvironmentConfig.setAllowCreate(true) set.
<br><br>Thanks very much for your help again John/George, I'll have a look at
<br>the documentation and try the setVerboseRecovery flag to see what
<br>messages are produced.
<br><br>Thanks,
<br>Ant
<br><br>-----Original Message-----
<br>From: George Feinberg [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5633572&i=0" target="_top" rel="nofollow">george.feinberg@...</a>]
<br>Sent: 03 August 2006 14:39
<br>To: Antony Grinyer
<br>Cc: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5633572&i=1" target="_top" rel="nofollow">xml@...</a>
<br>Subject: Re: [bdbxml] Unexpected Tomcat closure breaks containers?
<br><br>Antony,
<br><br>Are you also setting EnvironmentConfig.setAllowCreate(true) ?
<br>That's also necessary for recovery to be automatic.
<br>It's also necessary that recovery be run single-threaded, while
<br>no other threads or processes are using the environment.
<br><br>For debugging, another thing you can set is:
<br>EnvironmentConfig.setVerboseRecovery(true);
<br><br>and you'll see recovery-related messages during Environment
<br>creation.
<br><br>Regards,
<br><br>George
<br><div class='shrinkable-quote'><br>> Hi,
<br>>
<br>> Apologies to keep going on, but I am still unsure of a programmatic
<br>> solution to Tomcat crashing with containers open - we already set the
<br>> setRunRecovery(true) flag on the environment config which makes no
<br>> difference if Tomcat crashes and our containers have not been closed -
<br>> we're still not able to use our containers after we start Tomcat again
<br>> after a crash. Realistically can we only use the recovery shell
<br>> utility
<br>> to fix containers left open after a Tomcat crash?
<br>>
<br>> Thanks for your continued help,
<br>> Ant
<br>>
<br>> -----Original Message-----
<br>> From: John Snelson [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5633572&i=2" target="_top" rel="nofollow">john.snelson@...</a>]
<br>> Sent: 03 August 2006 11:55
<br>> To: Antony Grinyer
<br>> Cc: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5633572&i=3" target="_top" rel="nofollow">xml@...</a>
<br>> Subject: Re: [bdbxml] Unexpected Tomcat closure breaks containers?
<br>>
<br>> To run recovery using the java API, you simply need to to use
<br>> EnvironmentConfig.setRunRecovery(true) when you create the Environment
<br>> object:
<br>>
<br>> <a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/</a>
<br>> EnvironmentConfig
<br>> .html#setRunRecovery(boolean)
<br>>
<br>> Make sure that you follow the guidelines in the documentation
<br>> above, and
<br>>
<br>> run recovery in a single-threaded manner.
<br>>
<br>> DB XML will log to the Berkeley DB error stream or error handler,
<br>> which
<br>> is set using one of the following methods:
<br>>
<br>> <a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/</a>
<br>> EnvironmentConfig
<br>> .html#setErrorHandler(com.sleepycat.db.ErrorHandler)
<br>> <a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/</a>
<br>> EnvironmentConfig
<br>> .html#setErrorStream(java.io.OutputStream)
<br>>
<br>> To enable DB XML logging, you have to set the log categories and
<br>> levels
<br>> that you want to receive logging from:
<br>>
<br>> <a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/</a>
<br>> XmlManager.htm
<br>> l#setLogCategory(int,%20boolean)
<br>> <a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/</a>
<br>> XmlManager.htm
<br>> l#setLogLevel(int,%20boolean)
<br>>
<br>> Hope that helps,
<br>>
<br>> John
<br>>
<br>> Antony Grinyer wrote:
<br>>> Thanks for the help John; how do you run a recovery on a programmatic
<br>>> level in DB XML? From your advice, it would appear that running a
<br>>> recovery each time our application starts is most sensible, but
<br>>> how is
<br>>> this done in the DB XML API? We set the setRunRecovery=true
<br>> environment
<br>>> parameter but is there anything else we must do to ensure our
<br>> containers
<br>>> are recovered?
<br>>>
<br>>> On a slightly different note, does BD XML produce any of its own
<br>>> error
<br>>> logs we can read?
<br>>>
<br>>> Thanks again for your excellent support and advise.
<br>>>
<br>>> Ant
<br>>> ---
<br>>> Senior Technical Consultant
<br>>> CSW Group Ltd
<br>>>
<br>>> -----Original Message-----
<br>>> From: John Snelson [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5633572&i=4" target="_top" rel="nofollow">john.snelson@...</a>]
<br>>> Sent: 01 August 2006 18:08
<br>>> To: Antony Grinyer
<br>>> Cc: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5633572&i=5" target="_top" rel="nofollow">xml@...</a>
<br>>> Subject: Re: [bdbxml] Unexpected Tomcat closure breaks containers?
<br>>>
<br>>> Hi Antony,
<br>>>
<br>>> So long as you are using transactions, this situation is not a
<br>> problem.
<br>>> Since DB XML does not run as a separate server, the containers can
<br>> often
<br>>>
<br>>> be corrupted due to an application crash. In these situations, you
<br>> will
<br>>> need to run the recovery process, which will replay the transaction
<br>>> logs. This webpage should help you out:
<br>>>
<br>>> <a href="http://www.sleepycat.com/docs/ref/transapp/recovery.html" target="_top" rel="nofollow">http://www.sleepycat.com/docs/ref/transapp/recovery.html</a><br>>>
<br>>> Many people run recovery every time their applications start, since
<br>>> there is no penalty to running recovery when it is not needed.
<br>>>
<br>>> You should also be checkpointing, if you aren't already. This will
<br>>> reduce the amount of transaction logging that will need to be
<br>> replayed:
<br>>>
<br>>> <a href="http://www.sleepycat.com/docs/ref/transapp/checkpoint.html" target="_top" rel="nofollow">http://www.sleepycat.com/docs/ref/transapp/checkpoint.html</a><br>>>
<br>>> Hope that helps,
<br>>>
<br>>> John
<br>>>
<br>>> Antony Grinyer wrote:
<br>>>> Hi,
<br>>>>
<br>>>> We are using BDB XML 2.2.13 on Windows XP with Java 1.5. We use
<br>> Tomcat
<br>>>> for our web application and during start-up we open our containers
<br>>> using
<br>>>> a listener; we also close our containers using a listener when
<br>>>> Tomcat
<br>>>> shuts down, we do not use the container 'auto-open' environment
<br>>>> parameter. During a recent demo, Tomcat crashed unexpectedly which
<br>>>> appeared to corrupt all of our containers when we started Tomcat
<br>>> again.
<br>>>> Therefore, my question is what is the solution to ensure containers
<br>>> are
<br>>>> kept uncorrupted given a Tomcat crash? Is there any 'auto-close on
<br>>>> crash' type parameters in BDB XML? Has anyone implemented a fail
<br>>>> safe
<br>>>> scenario using Tomcat and BDB XML given a Tomcat crash?
<br>>>>
<br>>>> Thanks in advance for any help,
<br>>>>
<br>>>> Ant
<br>>>> ---
<br>>>> Senior Technical Consultant
<br>>>> CSW Group Ltd.
<br>>>>
<br>>>>
<br>>>>
<br>>>> ------------------------------------------
<br>>>> To remove yourself from this list, send an
<br>>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5633572&i=6" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>>
<br>>>
<br>>>
<br>>>
<br>>>
<br>>>
<br>>> ------------------------------------------
<br>>> To remove yourself from this list, send an
<br>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5633572&i=7" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>
<br>>
<br>>
<br>>
<br>>
<br>>
<br>> ------------------------------------------
<br>> To remove yourself from this list, send an
<br>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5633572&i=8" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>
</div><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5633572&i=9" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5632740Re: Unexpected Tomcat closure breaks containers?2006-08-03T07:39:28Z2006-08-03T07:39:28ZJohn Snelson-4The recovery shell program, db_recover, will do exactly the same as
<br>specifying setRunRecovery(true) when you open your Environment object.
<br><br>If you need more help, there is getting started with transactions guide
<br>that is available here:
<br><br><a href="http://www.sleepycat.com/docs/gsg_txn/JAVA/index.html" target="_top" rel="nofollow">http://www.sleepycat.com/docs/gsg_txn/JAVA/index.html</a><br><br>This should answer most of your questions.
<br><br>John
<br><br>Antony Grinyer wrote:
<div class='shrinkable-quote'><br>> Hi,
<br>>
<br>> Apologies to keep going on, but I am still unsure of a programmatic
<br>> solution to Tomcat crashing with containers open - we already set the
<br>> setRunRecovery(true) flag on the environment config which makes no
<br>> difference if Tomcat crashes and our containers have not been closed -
<br>> we're still not able to use our containers after we start Tomcat again
<br>> after a crash. Realistically can we only use the recovery shell utility
<br>> to fix containers left open after a Tomcat crash?
<br>>
<br>> Thanks for your continued help,
<br>> Ant
<br>>
<br>> -----Original Message-----
<br>> From: John Snelson [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632740&i=0" target="_top" rel="nofollow">john.snelson@...</a>]
<br>> Sent: 03 August 2006 11:55
<br>> To: Antony Grinyer
<br>> Cc: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632740&i=1" target="_top" rel="nofollow">xml@...</a>
<br>> Subject: Re: [bdbxml] Unexpected Tomcat closure breaks containers?
<br>>
<br>> To run recovery using the java API, you simply need to to use
<br>> EnvironmentConfig.setRunRecovery(true) when you create the Environment
<br>> object:
<br>>
<br>> <a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig</a><br>> .html#setRunRecovery(boolean)
<br>>
<br>> Make sure that you follow the guidelines in the documentation above, and
<br>>
<br>> run recovery in a single-threaded manner.
<br>>
<br>> DB XML will log to the Berkeley DB error stream or error handler, which
<br>> is set using one of the following methods:
<br>>
<br>> <a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig</a><br>> .html#setErrorHandler(com.sleepycat.db.ErrorHandler)
<br>> <a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig</a><br>> .html#setErrorStream(java.io.OutputStream)
<br>>
<br>> To enable DB XML logging, you have to set the log categories and levels
<br>> that you want to receive logging from:
<br>>
<br>> <a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/XmlManager.htm" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/XmlManager.htm</a><br>> l#setLogCategory(int,%20boolean)
<br>> <a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/XmlManager.htm" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/XmlManager.htm</a><br>> l#setLogLevel(int,%20boolean)
<br>>
<br>> Hope that helps,
<br>>
<br>> John
<br>>
<br>> Antony Grinyer wrote:
<br>>> Thanks for the help John; how do you run a recovery on a programmatic
<br>>> level in DB XML? From your advice, it would appear that running a
<br>>> recovery each time our application starts is most sensible, but how is
<br>>> this done in the DB XML API? We set the setRunRecovery=true
<br>> environment
<br>>> parameter but is there anything else we must do to ensure our
<br>> containers
<br>>> are recovered?
<br>>>
<br>>> On a slightly different note, does BD XML produce any of its own error
<br>>> logs we can read?
<br>>>
<br>>> Thanks again for your excellent support and advise.
<br>>>
<br>>> Ant
<br>>> ---
<br>>> Senior Technical Consultant
<br>>> CSW Group Ltd
<br>>>
<br>>> -----Original Message-----
<br>>> From: John Snelson [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632740&i=2" target="_top" rel="nofollow">john.snelson@...</a>]
<br>>> Sent: 01 August 2006 18:08
<br>>> To: Antony Grinyer
<br>>> Cc: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632740&i=3" target="_top" rel="nofollow">xml@...</a>
<br>>> Subject: Re: [bdbxml] Unexpected Tomcat closure breaks containers?
<br>>>
<br>>> Hi Antony,
<br>>>
<br>>> So long as you are using transactions, this situation is not a
<br>> problem.
<br>>> Since DB XML does not run as a separate server, the containers can
<br>> often
<br>>> be corrupted due to an application crash. In these situations, you
<br>> will
<br>>> need to run the recovery process, which will replay the transaction
<br>>> logs. This webpage should help you out:
<br>>>
<br>>> <a href="http://www.sleepycat.com/docs/ref/transapp/recovery.html" target="_top" rel="nofollow">http://www.sleepycat.com/docs/ref/transapp/recovery.html</a><br>>>
<br>>> Many people run recovery every time their applications start, since
<br>>> there is no penalty to running recovery when it is not needed.
<br>>>
<br>>> You should also be checkpointing, if you aren't already. This will
<br>>> reduce the amount of transaction logging that will need to be
<br>> replayed:
<br>>> <a href="http://www.sleepycat.com/docs/ref/transapp/checkpoint.html" target="_top" rel="nofollow">http://www.sleepycat.com/docs/ref/transapp/checkpoint.html</a><br>>>
<br>>> Hope that helps,
<br>>>
<br>>> John
<br>>>
<br>>> Antony Grinyer wrote:
<br>>>> Hi,
<br>>>>
<br>>>> We are using BDB XML 2.2.13 on Windows XP with Java 1.5. We use
<br>> Tomcat
<br>>>> for our web application and during start-up we open our containers
<br>>> using
<br>>>> a listener; we also close our containers using a listener when Tomcat
<br>>>> shuts down, we do not use the container 'auto-open' environment
<br>>>> parameter. During a recent demo, Tomcat crashed unexpectedly which
<br>>>> appeared to corrupt all of our containers when we started Tomcat
<br>>> again.
<br>>>> Therefore, my question is what is the solution to ensure containers
<br>>> are
<br>>>> kept uncorrupted given a Tomcat crash? Is there any 'auto-close on
<br>>>> crash' type parameters in BDB XML? Has anyone implemented a fail safe
<br>>>> scenario using Tomcat and BDB XML given a Tomcat crash?
<br>>>>
<br>>>> Thanks in advance for any help,
<br>>>>
<br>>>> Ant
<br>>>> ---
<br>>>> Senior Technical Consultant
<br>>>> CSW Group Ltd.
<br>>>>
<br>>>>
<br>>>>
<br>>>> ------------------------------------------
<br>>>> To remove yourself from this list, send an
<br>>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632740&i=4" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>>
<br>>>
<br>>>
<br>>>
<br>>>
<br>>> ------------------------------------------
<br>>> To remove yourself from this list, send an
<br>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632740&i=5" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>
<br>>
<br>>
<br>>
<br>>
<br>>
<br>> ------------------------------------------
<br>> To remove yourself from this list, send an
<br>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632740&i=6" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>
</div><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632740&i=7" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5632873Re: Unexpected Tomcat closure breaks containers?2006-08-03T07:39:23Z2006-08-03T07:39:23ZGeorge Feinberg-3Antony,
<br><br>Are you also setting EnvironmentConfig.setAllowCreate(true) ?
<br>That's also necessary for recovery to be automatic.
<br>It's also necessary that recovery be run single-threaded, while
<br>no other threads or processes are using the environment.
<br><br>For debugging, another thing you can set is:
<br>EnvironmentConfig.setVerboseRecovery(true);
<br><br>and you'll see recovery-related messages during Environment
<br>creation.
<br><br>Regards,
<br><br>George
<br><div class='shrinkable-quote'><br>> Hi,
<br>>
<br>> Apologies to keep going on, but I am still unsure of a programmatic
<br>> solution to Tomcat crashing with containers open - we already set the
<br>> setRunRecovery(true) flag on the environment config which makes no
<br>> difference if Tomcat crashes and our containers have not been closed -
<br>> we're still not able to use our containers after we start Tomcat again
<br>> after a crash. Realistically can we only use the recovery shell
<br>> utility
<br>> to fix containers left open after a Tomcat crash?
<br>>
<br>> Thanks for your continued help,
<br>> Ant
<br>>
<br>> -----Original Message-----
<br>> From: John Snelson [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632873&i=0" target="_top" rel="nofollow">john.snelson@...</a>]
<br>> Sent: 03 August 2006 11:55
<br>> To: Antony Grinyer
<br>> Cc: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632873&i=1" target="_top" rel="nofollow">xml@...</a>
<br>> Subject: Re: [bdbxml] Unexpected Tomcat closure breaks containers?
<br>>
<br>> To run recovery using the java API, you simply need to to use
<br>> EnvironmentConfig.setRunRecovery(true) when you create the Environment
<br>> object:
<br>>
<br>> <a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/</a>
<br>> EnvironmentConfig
<br>> .html#setRunRecovery(boolean)
<br>>
<br>> Make sure that you follow the guidelines in the documentation
<br>> above, and
<br>>
<br>> run recovery in a single-threaded manner.
<br>>
<br>> DB XML will log to the Berkeley DB error stream or error handler,
<br>> which
<br>> is set using one of the following methods:
<br>>
<br>> <a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/</a>
<br>> EnvironmentConfig
<br>> .html#setErrorHandler(com.sleepycat.db.ErrorHandler)
<br>> <a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/</a>
<br>> EnvironmentConfig
<br>> .html#setErrorStream(java.io.OutputStream)
<br>>
<br>> To enable DB XML logging, you have to set the log categories and
<br>> levels
<br>> that you want to receive logging from:
<br>>
<br>> <a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/</a>
<br>> XmlManager.htm
<br>> l#setLogCategory(int,%20boolean)
<br>> <a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/</a>
<br>> XmlManager.htm
<br>> l#setLogLevel(int,%20boolean)
<br>>
<br>> Hope that helps,
<br>>
<br>> John
<br>>
<br>> Antony Grinyer wrote:
<br>>> Thanks for the help John; how do you run a recovery on a programmatic
<br>>> level in DB XML? From your advice, it would appear that running a
<br>>> recovery each time our application starts is most sensible, but
<br>>> how is
<br>>> this done in the DB XML API? We set the setRunRecovery=true
<br>> environment
<br>>> parameter but is there anything else we must do to ensure our
<br>> containers
<br>>> are recovered?
<br>>>
<br>>> On a slightly different note, does BD XML produce any of its own
<br>>> error
<br>>> logs we can read?
<br>>>
<br>>> Thanks again for your excellent support and advise.
<br>>>
<br>>> Ant
<br>>> ---
<br>>> Senior Technical Consultant
<br>>> CSW Group Ltd
<br>>>
<br>>> -----Original Message-----
<br>>> From: John Snelson [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632873&i=2" target="_top" rel="nofollow">john.snelson@...</a>]
<br>>> Sent: 01 August 2006 18:08
<br>>> To: Antony Grinyer
<br>>> Cc: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632873&i=3" target="_top" rel="nofollow">xml@...</a>
<br>>> Subject: Re: [bdbxml] Unexpected Tomcat closure breaks containers?
<br>>>
<br>>> Hi Antony,
<br>>>
<br>>> So long as you are using transactions, this situation is not a
<br>> problem.
<br>>> Since DB XML does not run as a separate server, the containers can
<br>> often
<br>>>
<br>>> be corrupted due to an application crash. In these situations, you
<br>> will
<br>>> need to run the recovery process, which will replay the transaction
<br>>> logs. This webpage should help you out:
<br>>>
<br>>> <a href="http://www.sleepycat.com/docs/ref/transapp/recovery.html" target="_top" rel="nofollow">http://www.sleepycat.com/docs/ref/transapp/recovery.html</a><br>>>
<br>>> Many people run recovery every time their applications start, since
<br>>> there is no penalty to running recovery when it is not needed.
<br>>>
<br>>> You should also be checkpointing, if you aren't already. This will
<br>>> reduce the amount of transaction logging that will need to be
<br>> replayed:
<br>>>
<br>>> <a href="http://www.sleepycat.com/docs/ref/transapp/checkpoint.html" target="_top" rel="nofollow">http://www.sleepycat.com/docs/ref/transapp/checkpoint.html</a><br>>>
<br>>> Hope that helps,
<br>>>
<br>>> John
<br>>>
<br>>> Antony Grinyer wrote:
<br>>>> Hi,
<br>>>>
<br>>>> We are using BDB XML 2.2.13 on Windows XP with Java 1.5. We use
<br>> Tomcat
<br>>>> for our web application and during start-up we open our containers
<br>>> using
<br>>>> a listener; we also close our containers using a listener when
<br>>>> Tomcat
<br>>>> shuts down, we do not use the container 'auto-open' environment
<br>>>> parameter. During a recent demo, Tomcat crashed unexpectedly which
<br>>>> appeared to corrupt all of our containers when we started Tomcat
<br>>> again.
<br>>>> Therefore, my question is what is the solution to ensure containers
<br>>> are
<br>>>> kept uncorrupted given a Tomcat crash? Is there any 'auto-close on
<br>>>> crash' type parameters in BDB XML? Has anyone implemented a fail
<br>>>> safe
<br>>>> scenario using Tomcat and BDB XML given a Tomcat crash?
<br>>>>
<br>>>> Thanks in advance for any help,
<br>>>>
<br>>>> Ant
<br>>>> ---
<br>>>> Senior Technical Consultant
<br>>>> CSW Group Ltd.
<br>>>>
<br>>>>
<br>>>>
<br>>>> ------------------------------------------
<br>>>> To remove yourself from this list, send an
<br>>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632873&i=4" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>>
<br>>>
<br>>>
<br>>>
<br>>>
<br>>>
<br>>> ------------------------------------------
<br>>> To remove yourself from this list, send an
<br>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632873&i=5" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>
<br>>
<br>>
<br>>
<br>>
<br>>
<br>> ------------------------------------------
<br>> To remove yourself from this list, send an
<br>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632873&i=6" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>
</div><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632873&i=7" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5632341RE: Unexpected Tomcat closure breaks containers?2006-08-03T07:16:58Z2006-08-03T07:16:58ZAntony GrinyerHi,
<br><br>Apologies to keep going on, but I am still unsure of a programmatic
<br>solution to Tomcat crashing with containers open - we already set the
<br>setRunRecovery(true) flag on the environment config which makes no
<br>difference if Tomcat crashes and our containers have not been closed -
<br>we're still not able to use our containers after we start Tomcat again
<br>after a crash. Realistically can we only use the recovery shell utility
<br>to fix containers left open after a Tomcat crash?
<br><br>Thanks for your continued help,
<br>Ant
<br><br>-----Original Message-----
<br>From: John Snelson [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632341&i=0" target="_top" rel="nofollow">john.snelson@...</a>]
<br>Sent: 03 August 2006 11:55
<br>To: Antony Grinyer
<br>Cc: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632341&i=1" target="_top" rel="nofollow">xml@...</a>
<br>Subject: Re: [bdbxml] Unexpected Tomcat closure breaks containers?
<br><br>To run recovery using the java API, you simply need to to use
<br>EnvironmentConfig.setRunRecovery(true) when you create the Environment
<br>object:
<br><br><a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig</a><br>.html#setRunRecovery(boolean)
<br><br>Make sure that you follow the guidelines in the documentation above, and
<br><br>run recovery in a single-threaded manner.
<br><br>DB XML will log to the Berkeley DB error stream or error handler, which
<br>is set using one of the following methods:
<br><br><a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig</a><br>.html#setErrorHandler(com.sleepycat.db.ErrorHandler)
<br><a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig</a><br>.html#setErrorStream(java.io.OutputStream)
<br><br>To enable DB XML logging, you have to set the log categories and levels
<br>that you want to receive logging from:
<br><br><a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/XmlManager.htm" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/XmlManager.htm</a><br>l#setLogCategory(int,%20boolean)
<br><a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/XmlManager.htm" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/XmlManager.htm</a><br>l#setLogLevel(int,%20boolean)
<br><br>Hope that helps,
<br><br>John
<br><br>Antony Grinyer wrote:
<br>> Thanks for the help John; how do you run a recovery on a programmatic
<br>> level in DB XML? From your advice, it would appear that running a
<br>> recovery each time our application starts is most sensible, but how is
<br>> this done in the DB XML API? We set the setRunRecovery=true
<br>environment
<br>> parameter but is there anything else we must do to ensure our
<br>containers
<div class='shrinkable-quote'><br>> are recovered?
<br>>
<br>> On a slightly different note, does BD XML produce any of its own error
<br>> logs we can read?
<br>>
<br>> Thanks again for your excellent support and advise.
<br>>
<br>> Ant
<br>> ---
<br>> Senior Technical Consultant
<br>> CSW Group Ltd
<br>>
<br>> -----Original Message-----
<br>> From: John Snelson [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632341&i=2" target="_top" rel="nofollow">john.snelson@...</a>]
<br>> Sent: 01 August 2006 18:08
<br>> To: Antony Grinyer
<br>> Cc: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632341&i=3" target="_top" rel="nofollow">xml@...</a>
<br>> Subject: Re: [bdbxml] Unexpected Tomcat closure breaks containers?
<br>>
<br>> Hi Antony,
<br>>
<br>> So long as you are using transactions, this situation is not a
</div>problem.
<br>> Since DB XML does not run as a separate server, the containers can
<br>often
<br>>
<br>> be corrupted due to an application crash. In these situations, you
<br>will
<div class='shrinkable-quote'><br>> need to run the recovery process, which will replay the transaction
<br>> logs. This webpage should help you out:
<br>>
<br>> <a href="http://www.sleepycat.com/docs/ref/transapp/recovery.html" target="_top" rel="nofollow">http://www.sleepycat.com/docs/ref/transapp/recovery.html</a><br>>
<br>> Many people run recovery every time their applications start, since
<br>> there is no penalty to running recovery when it is not needed.
<br>>
<br>> You should also be checkpointing, if you aren't already. This will
<br>> reduce the amount of transaction logging that will need to be
</div>replayed:
<div class='shrinkable-quote'><br>>
<br>> <a href="http://www.sleepycat.com/docs/ref/transapp/checkpoint.html" target="_top" rel="nofollow">http://www.sleepycat.com/docs/ref/transapp/checkpoint.html</a><br>>
<br>> Hope that helps,
<br>>
<br>> John
<br>>
<br>> Antony Grinyer wrote:
<br>>> Hi,
<br>>>
<br>>> We are using BDB XML 2.2.13 on Windows XP with Java 1.5. We use
</div>Tomcat
<div class='shrinkable-quote'><br>>> for our web application and during start-up we open our containers
<br>> using
<br>>> a listener; we also close our containers using a listener when Tomcat
<br>>> shuts down, we do not use the container 'auto-open' environment
<br>>> parameter. During a recent demo, Tomcat crashed unexpectedly which
<br>>> appeared to corrupt all of our containers when we started Tomcat
<br>> again.
<br>>> Therefore, my question is what is the solution to ensure containers
<br>> are
<br>>> kept uncorrupted given a Tomcat crash? Is there any 'auto-close on
<br>>> crash' type parameters in BDB XML? Has anyone implemented a fail safe
<br>>> scenario using Tomcat and BDB XML given a Tomcat crash?
<br>>>
<br>>> Thanks in advance for any help,
<br>>>
<br>>> Ant
<br>>> ---
<br>>> Senior Technical Consultant
<br>>> CSW Group Ltd.
<br>>>
<br>>>
<br>>>
<br>>> ------------------------------------------
<br>>> To remove yourself from this list, send an
<br>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632341&i=4" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>
<br>>
<br>>
<br>>
<br>>
<br>>
<br>> ------------------------------------------
<br>> To remove yourself from this list, send an
<br>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632341&i=5" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>
</div><br><br><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5632341&i=6" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5630413Re: Unexpected Tomcat closure breaks containers?2006-08-03T04:55:27Z2006-08-03T04:55:27ZJohn Snelson-4To run recovery using the java API, you simply need to to use
<br>EnvironmentConfig.setRunRecovery(true) when you create the Environment
<br>object:
<br><br><a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig.html#setRunRecovery(boolean" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig.html#setRunRecovery(boolean</a>)
<br><br>Make sure that you follow the guidelines in the documentation above, and
<br>run recovery in a single-threaded manner.
<br><br>DB XML will log to the Berkeley DB error stream or error handler, which
<br>is set using one of the following methods:
<br><br><a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig.html#setErrorHandler(com.sleepycat.db.ErrorHandler" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig.html#setErrorHandler(com.sleepycat.db.ErrorHandler</a>)
<br><a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig.html#setErrorStream(java.io.OutputStream" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/db/EnvironmentConfig.html#setErrorStream(java.io.OutputStream</a>)
<br><br>To enable DB XML logging, you have to set the log categories and levels
<br>that you want to receive logging from:
<br><br><a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/XmlManager.html#setLogCategory(int,%20boolean" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/XmlManager.html#setLogCategory(int,%20boolean</a>)
<br><a href="http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/XmlManager.html#setLogLevel(int,%20boolean" target="_top" rel="nofollow">http://www.sleepycat.com/xmldocs/java/com/sleepycat/dbxml/XmlManager.html#setLogLevel(int,%20boolean</a>)
<br><br>Hope that helps,
<br><br>John
<br><br>Antony Grinyer wrote:
<div class='shrinkable-quote'><br>> Thanks for the help John; how do you run a recovery on a programmatic
<br>> level in DB XML? From your advice, it would appear that running a
<br>> recovery each time our application starts is most sensible, but how is
<br>> this done in the DB XML API? We set the setRunRecovery=true environment
<br>> parameter but is there anything else we must do to ensure our containers
<br>> are recovered?
<br>>
<br>> On a slightly different note, does BD XML produce any of its own error
<br>> logs we can read?
<br>>
<br>> Thanks again for your excellent support and advise.
<br>>
<br>> Ant
<br>> ---
<br>> Senior Technical Consultant
<br>> CSW Group Ltd
<br>>
<br>> -----Original Message-----
<br>> From: John Snelson [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5630413&i=0" target="_top" rel="nofollow">john.snelson@...</a>]
<br>> Sent: 01 August 2006 18:08
<br>> To: Antony Grinyer
<br>> Cc: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5630413&i=1" target="_top" rel="nofollow">xml@...</a>
<br>> Subject: Re: [bdbxml] Unexpected Tomcat closure breaks containers?
<br>>
<br>> Hi Antony,
<br>>
<br>> So long as you are using transactions, this situation is not a problem.
<br>> Since DB XML does not run as a separate server, the containers can often
<br>>
<br>> be corrupted due to an application crash. In these situations, you will
<br>> need to run the recovery process, which will replay the transaction
<br>> logs. This webpage should help you out:
<br>>
<br>> <a href="http://www.sleepycat.com/docs/ref/transapp/recovery.html" target="_top" rel="nofollow">http://www.sleepycat.com/docs/ref/transapp/recovery.html</a><br>>
<br>> Many people run recovery every time their applications start, since
<br>> there is no penalty to running recovery when it is not needed.
<br>>
<br>> You should also be checkpointing, if you aren't already. This will
<br>> reduce the amount of transaction logging that will need to be replayed:
<br>>
<br>> <a href="http://www.sleepycat.com/docs/ref/transapp/checkpoint.html" target="_top" rel="nofollow">http://www.sleepycat.com/docs/ref/transapp/checkpoint.html</a><br>>
<br>> Hope that helps,
<br>>
<br>> John
<br>>
<br>> Antony Grinyer wrote:
<br>>> Hi,
<br>>>
<br>>> We are using BDB XML 2.2.13 on Windows XP with Java 1.5. We use Tomcat
<br>>> for our web application and during start-up we open our containers
<br>> using
<br>>> a listener; we also close our containers using a listener when Tomcat
<br>>> shuts down, we do not use the container 'auto-open' environment
<br>>> parameter. During a recent demo, Tomcat crashed unexpectedly which
<br>>> appeared to corrupt all of our containers when we started Tomcat
<br>> again.
<br>>> Therefore, my question is what is the solution to ensure containers
<br>> are
<br>>> kept uncorrupted given a Tomcat crash? Is there any 'auto-close on
<br>>> crash' type parameters in BDB XML? Has anyone implemented a fail safe
<br>>> scenario using Tomcat and BDB XML given a Tomcat crash?
<br>>>
<br>>> Thanks in advance for any help,
<br>>>
<br>>> Ant
<br>>> ---
<br>>> Senior Technical Consultant
<br>>> CSW Group Ltd.
<br>>>
<br>>>
<br>>>
<br>>> ------------------------------------------
<br>>> To remove yourself from this list, send an
<br>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5630413&i=2" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>
<br>>
<br>>
<br>>
<br>>
<br>>
<br>> ------------------------------------------
<br>> To remove yourself from this list, send an
<br>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5630413&i=3" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>
</div><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5630413&i=4" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5629010RE: Unexpected Tomcat closure breaks containers?2006-08-03T02:54:08Z2006-08-03T02:54:08ZAntony GrinyerThanks for the help John; how do you run a recovery on a programmatic
<br>level in DB XML? From your advice, it would appear that running a
<br>recovery each time our application starts is most sensible, but how is
<br>this done in the DB XML API? We set the setRunRecovery=true environment
<br>parameter but is there anything else we must do to ensure our containers
<br>are recovered?
<br><br>On a slightly different note, does BD XML produce any of its own error
<br>logs we can read?
<br><br>Thanks again for your excellent support and advise.
<br><br>Ant
<br>---
<br>Senior Technical Consultant
<br>CSW Group Ltd
<br><br>-----Original Message-----
<br>From: John Snelson [mailto:<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5629010&i=0" target="_top" rel="nofollow">john.snelson@...</a>]
<br>Sent: 01 August 2006 18:08
<br>To: Antony Grinyer
<br>Cc: <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5629010&i=1" target="_top" rel="nofollow">xml@...</a>
<br>Subject: Re: [bdbxml] Unexpected Tomcat closure breaks containers?
<br><br>Hi Antony,
<br><br>So long as you are using transactions, this situation is not a problem.
<br>Since DB XML does not run as a separate server, the containers can often
<br><br>be corrupted due to an application crash. In these situations, you will
<br>need to run the recovery process, which will replay the transaction
<br>logs. This webpage should help you out:
<br><br><a href="http://www.sleepycat.com/docs/ref/transapp/recovery.html" target="_top" rel="nofollow">http://www.sleepycat.com/docs/ref/transapp/recovery.html</a><br><br>Many people run recovery every time their applications start, since
<br>there is no penalty to running recovery when it is not needed.
<br><br>You should also be checkpointing, if you aren't already. This will
<br>reduce the amount of transaction logging that will need to be replayed:
<br><br><a href="http://www.sleepycat.com/docs/ref/transapp/checkpoint.html" target="_top" rel="nofollow">http://www.sleepycat.com/docs/ref/transapp/checkpoint.html</a><br><br>Hope that helps,
<br><br>John
<br><br>Antony Grinyer wrote:
<br>> Hi,
<br>>
<br>> We are using BDB XML 2.2.13 on Windows XP with Java 1.5. We use Tomcat
<br>> for our web application and during start-up we open our containers
<br>using
<br>> a listener; we also close our containers using a listener when Tomcat
<br>> shuts down, we do not use the container 'auto-open' environment
<br>> parameter. During a recent demo, Tomcat crashed unexpectedly which
<br>> appeared to corrupt all of our containers when we started Tomcat
<br>again.
<br>> Therefore, my question is what is the solution to ensure containers
<br>are
<div class='shrinkable-quote'><br>> kept uncorrupted given a Tomcat crash? Is there any 'auto-close on
<br>> crash' type parameters in BDB XML? Has anyone implemented a fail safe
<br>> scenario using Tomcat and BDB XML given a Tomcat crash?
<br>>
<br>> Thanks in advance for any help,
<br>>
<br>> Ant
<br>> ---
<br>> Senior Technical Consultant
<br>> CSW Group Ltd.
<br>>
<br>>
<br>>
<br>> ------------------------------------------
<br>> To remove yourself from this list, send an
<br>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5629010&i=2" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>
</div><br><br><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5629010&i=3" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5616921Re: Reminder: Berkeley DB XML mailing list transition2006-08-02T09:55:27Z2006-08-02T09:55:27ZGeorge Feinberg-3John, all,
<br><br>We welcome feedback on this issue. It would be a bad thing to
<br>lose even a small portion of the BDB XML community that has been
<br>built by those on this list. The development team very much
<br>appreciates the excellent user participation on this mailing list,
<br>and that feedback makes the product what it is.
<br><br>It will be unfortunate if we lose people in the planned transition.
<br>Please
<br>be patient, and persistent! We will work within our new home at
<br>Oracle to keep
<br>the BDB XML community together and happy.
<br><br>If anyone has comments, suggestions or advice about how to keep
<br>people involved, and make participation easier, we would love to hear
<br>from you,
<br>publicly or privately. Even if it's just to reiterate issues that
<br>have been brought up.
<br><br>Thanks for your support!
<br><br>Regards,
<br><br>George, John, and the BDB XML team
<br><br><br>On Aug 2, 2006, at 12:22 AM, John Ralls wrote:
<br><div class='shrinkable-quote'><br>> On Aug 1, 2006, at 8:01 AM, George Feinberg wrote:
<br>>
<br>>> A reminder. Traffic on this mailing list should move to the
<br>>> Berkeley DB XML forum:
<br>>> <a href="http://forums.oracle.com/forums/forum.jspa?forumID=274" target="_top" rel="nofollow">http://forums.oracle.com/forums/forum.jspa?forumID=274</a><br>>>
<br>>> All of the Sleepycat Software Berkeley DB support channels will
<br>>> move to Oracle support channels
<br>>> during the month of August.
<br>>>
<br>>> This includes the Berkeley DB Java Edition, Berkeley DB XML and
<br>>> Berkeley DB High Availability product mailing lists.
<br>>>
<br>>> The mailing lists are being moved to the Oracle Technical Network
<br>>> forums:
<br>>> <a href="http://forums.oracle.com/forums/category.jspa?categoryID=18" target="_top" rel="nofollow">http://forums.oracle.com/forums/category.jspa?categoryID=18</a><br>>>
<br>>> There are four technical forums at this site: one each for
<br>>> Berkeley DB, Berkeley DB High Availability, Berkeley DB XML and
<br>>> Berkeley DB Java Edition. Please select the appropriate forum and
<br>>> post your question there.
<br>>>
<br>>> The Berkeley DB support and engineering teams are already
<br>>> monitoring these forums, and you can expect prompt and helpful
<br>>> answers.
<br>>>
<br>>> After an appropriate transition period, the Berkeley DB mailing
<br>>> lists will be turned off.
<br>>>
<br>>> While it will no longer be possible to send email directly to a
<br>>> mailing
<br>>> list, the use of an OTN forum has a couple of features that you
<br>>> may appreciate:
<br>>> 1. The forum is searchable
<br>>> 2. An RSS feed is available
<br>>> 3. You can "watch" a forum, getting email notifications of new
<br>>> postings,
<br>>> with the content of the post.
<br>>> 4. You can also watch specific threads or users.
<br>>> 5. You can read, and search the forum as "guest" but you need to
<br>>> register
<br>>> and sign in to post.
<br>>>
<br>>> Regards,
<br>>>
<br>>> George Feinberg
<br>>
<br>> Well, the RSS is pretty much useless. It transmits the message only
<br>> to the first \n.
<br>> The email from the watch function is seriously ugly: It also
<br>> appears that one can't post via email, which makes it largely
<br>> useful only for lurking.
<br>> And registration is clearly oriented toward corporate accounts
<br>> rather than the open source community, with lots of impertinent
<br>> "required" fields which are none of Oracle's business. (Required to
<br>> be filled in, of course. No requirement for truth. ;-) )
<br>>
<br>> Good way to ensure that the only support is from "official channels".
<br>>
<br>> Regards,
<br>> John Ralls
<br>>
<br>>
<br>>
<br>> ------------------------------------------
<br>> To remove yourself from this list, send an
<br>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5616921&i=0" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>
</div><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5616921&i=1" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5607582Re: Reminder: Berkeley DB XML mailing list transition2006-08-01T22:22:00Z2006-08-01T22:22:00ZJohn Ralls<br>On Aug 1, 2006, at 8:01 AM, George Feinberg wrote:
<br><div class='shrinkable-quote'><br>> A reminder. Traffic on this mailing list should move to the
<br>> Berkeley DB XML forum:
<br>> <a href="http://forums.oracle.com/forums/forum.jspa?forumID=274" target="_top" rel="nofollow">http://forums.oracle.com/forums/forum.jspa?forumID=274</a><br>>
<br>> All of the Sleepycat Software Berkeley DB support channels will
<br>> move to Oracle support channels
<br>> during the month of August.
<br>>
<br>> This includes the Berkeley DB Java Edition, Berkeley DB XML and
<br>> Berkeley DB High Availability product mailing lists.
<br>>
<br>> The mailing lists are being moved to the Oracle Technical Network
<br>> forums:
<br>> <a href="http://forums.oracle.com/forums/category.jspa?categoryID=18" target="_top" rel="nofollow">http://forums.oracle.com/forums/category.jspa?categoryID=18</a><br>>
<br>> There are four technical forums at this site: one each for Berkeley
<br>> DB, Berkeley DB High Availability, Berkeley DB XML and Berkeley DB
<br>> Java Edition. Please select the appropriate forum and post your
<br>> question there.
<br>>
<br>> The Berkeley DB support and engineering teams are already
<br>> monitoring these forums, and you can expect prompt and helpful
<br>> answers.
<br>>
<br>> After an appropriate transition period, the Berkeley DB mailing
<br>> lists will be turned off.
<br>>
<br>> While it will no longer be possible to send email directly to a
<br>> mailing
<br>> list, the use of an OTN forum has a couple of features that you may
<br>> appreciate:
<br>> 1. The forum is searchable
<br>> 2. An RSS feed is available
<br>> 3. You can "watch" a forum, getting email notifications of new
<br>> postings,
<br>> with the content of the post.
<br>> 4. You can also watch specific threads or users.
<br>> 5. You can read, and search the forum as "guest" but you need to
<br>> register
<br>> and sign in to post.
<br>>
<br>> Regards,
<br>>
<br>> George Feinberg
</div><br>Well, the RSS is pretty much useless. It transmits the message only
<br>to the first \n.
<br>The email from the watch function is seriously ugly: It also appears
<br>that one can't post via email, which makes it largely useful only for
<br>lurking.
<br>And registration is clearly oriented toward corporate accounts rather
<br>than the open source community, with lots of impertinent "required"
<br>fields which are none of Oracle's business. (Required to be filled
<br>in, of course. No requirement for truth. ;-) )
<br><br>Good way to ensure that the only support is from "official channels".
<br><br>Regards,
<br>John Ralls
<br><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5607582&i=0" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5604870Re: XQuery boolean datatype handling2006-08-01T17:04:30Z2006-08-01T17:04:30ZMatt Magoffin-4Sorry I hadn't noticed the post about moving to the Oracle forums... I
<br>will re-post this message there.
<br><br>-- m@
<br><div class='shrinkable-quote'><br>> If I run this XQuery:
<br>>
<br>> for $match in collection('matchback.dbxml')/als:match-back-matches
<br>> where $match/@has-match = true()
<br>> return $match
<br>>
<br>> I get a single result (which is not expected), while if I change true() to
<br>> 'true' like this:
<br>>
<br>> for $match in collection('matchback.dbxml')/als:match-back-matches
<br>> where $match/@has-match = 'true'
<br>> return $match
<br>>
<br>> I get many, many results (which is expected). In general I've seen this
<br>> type of boolean handling discrepancy in many of my queries. I would have
<br>> thought if the true() function was an inappropriate use in this query,
<br>> that no results would be returned, not "some" results.
<br>>
<br>> I also notice the optimized query plans are slightly different for these
<br>> queries:
<br>>
<br>> With true():
<br>> n(V(node-attribute-equality-boolean,@has-match,=,'true'),P(node-element-presence-none,=,match-back-matches:<a href="http://autoleadservice.com/xml/als" target="_top" rel="nofollow">http://autoleadservice.com/xml/als</a>))
<br>>
<br>> With 'true':
<br>> n(P(node-element-presence-none,=,match-back-matches:<a href="http://autoleadservice.com/xml/als),P(node-attribute-equality-boolean,prefix,@has-match" target="_top" rel="nofollow">http://autoleadservice.com/xml/als),P(node-attribute-equality-boolean,prefix,@has-match</a>))
<br>>
<br>> I have a boolean index defined on the @has-match attribute, as well a a
<br>> default string-equality index for all attributes. Is it OK to have
<br>> overlapping indices defined in this way?
<br>>
<br>> -- m@
<br>>
<br>>
<br>>
<br>> ------------------------------------------
<br>> To remove yourself from this list, send an
<br>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5604870&i=0" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>
<br>>
</div><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5604870&i=1" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5604806XQuery boolean datatype handling2006-08-01T16:58:32Z2006-08-01T16:58:32ZMatt Magoffin-4If I run this XQuery:
<br><br>for $match in collection('matchback.dbxml')/als:match-back-matches
<br>where $match/@has-match = true()
<br>return $match
<br><br>I get a single result (which is not expected), while if I change true() to
<br>'true' like this:
<br><br>for $match in collection('matchback.dbxml')/als:match-back-matches
<br>where $match/@has-match = 'true'
<br>return $match
<br><br>I get many, many results (which is expected). In general I've seen this
<br>type of boolean handling discrepancy in many of my queries. I would have
<br>thought if the true() function was an inappropriate use in this query,
<br>that no results would be returned, not "some" results.
<br><br>I also notice the optimized query plans are slightly different for these
<br>queries:
<br><br>With true():
<br>n(V(node-attribute-equality-boolean,@has-match,=,'true'),P(node-element-presence-none,=,match-back-matches:<a href="http://autoleadservice.com/xml/als" target="_top" rel="nofollow">http://autoleadservice.com/xml/als</a>))
<br><br>With 'true':
<br>n(P(node-element-presence-none,=,match-back-matches:<a href="http://autoleadservice.com/xml/als),P(node-attribute-equality-boolean,prefix,@has-match" target="_top" rel="nofollow">http://autoleadservice.com/xml/als),P(node-attribute-equality-boolean,prefix,@has-match</a>))
<br><br>I have a boolean index defined on the @has-match attribute, as well a a
<br>default string-equality index for all attributes. Is it OK to have
<br>overlapping indices defined in this way?
<br><br>-- m@
<br><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5604806&i=0" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5603782Re: DOCUMENT_NOT_FOUND when try to delete while iterating over XQuery result set2006-08-01T15:38:04Z2006-08-01T15:38:04ZMatt Magoffin-4I added a catch for this exception, and noticed for one query 117
<br>DOCUMENT_NOT_FOUND exceptions for only 2 actual deletions. For the 117
<br>exceptions, they all appear to have unique document names.
<br><br>This problem seems somewhat related to the other DOCUMENT_NOT_FOUND
<br>problem I noted another email thread... because after adding the check to
<br>ignore the error and continue, my application then attempted to load data
<br>in and ran into the DOCUMENT_NOT_FOUND during a updateDocument() call.
<br><br>-- m@
<br><div class='shrinkable-quote'><br>> Matt,
<br>>
<br>> The most obvious explanation is that you are getting more
<br>> than one result for that document, and removing it
<br>> based on the first result. Then, when you process
<br>> your second result, the document's gone.
<br>>
<br>> If this is possible, just catch, and ignore this exception.
<br>>
<br>> You can change your deleteDocument() call to one
<br>> that just prints out the document name to see if
<br>> this is happening.
<br>>
<br>> George
<br>>
<br>> On Jul 31, 2006, at 10:59 PM, Matt Magoffin wrote:
<br>>
<br>>> Sorry, actually I spoke too soon. I still see the below exception,
<br>>> even
<br>>> with DB 4.4.
<br>>>
<br>>> Any insight would be appreciated!
<br>>>
<br>>> -- m@
<br>>>
<br>>>> I recompiled BDB XML against DB 4.4.20, and the problems seem to
<br>>>> have gone
<br>>>> away. I had been using the DB 4.3 that came with XML 2.2.13
<br>>>> previously.
<br>>>>
<br>>>> -- m@
<br>>>>
<br>>>>> Hello,
<br>>>>>
<br>>>>> I've been having a problem appear every so often that I can't
<br>>>>> seem to
<br>>>>> track down. It seems to happen more frequently after loading sets of
<br>>>>> data,
<br>>>>> but I cannot be sure exactly of the cause.
<br>>>>>
<br>>>>> The symptom is that I execute an XQuery and then try to delete
<br>>>>> all the
<br>>>>> documents resulting from the XQuery, all within a transaction. If
<br>>>>> I step
<br>>>>> through the code in a debugger, I see BDB XML returning the
<br>>>>> XmlDocument
<br>>>>> instances, but then throws the below exception:
<br>>>>>
<br>>>>> -----
<br>>>>> Error: Document not found: 4991476, errcode = DOCUMENT_NOT_FOUND
<br>>>>> at
<br>>>>> com.sleepycat.dbxml.dbxml_javaJNI.XmlContainer_deleteDocument__SWIG_
<br>>>>> 2(Native
<br>>>>> Method)
<br>>>>> at
<br>>>>> com.sleepycat.dbxml.XmlContainer.deleteDocument(XmlContainer.java:
<br>>>>> 367)
<br>>>>> -----
<br>>>>>
<br>>>>> The general flow of the program is this:
<br>>>>>
<br>>>>> -----
<br>>>>> XmlQueryContext context = xmlManager.createQueryContext();
<br>>>>> XmlQueryExpression qe = createQueryExpression(criteria, context,
<br>>>>> XmlQueryContext.LiveValues, XmlQueryContext.Lazy);
<br>>>>> XmlResults results = qe.execute(holder.getTransaction(), context);
<br>>>>>
<br>>>>> XmlUpdateContext updateContext = xmlManager.createUpdateContext();
<br>>>>> XmlDocument docToDelete = xmlManager.createDocument();
<br>>>>> while ( results.next(docToDelete) ) {
<br>>>>> xmlContainer.deleteDocument(holder.getTransaction(),
<br>>>>> docToDelete, updateContext);
<br>>>>> }
<br>>>>> -----
<br>>>>>
<br>>>>> I've tried experimenting with XmlQueryContext.Eager, but that did
<br>>>>> not
<br>>>>> seem
<br>>>>> to make any difference.
<br>>>>>
<br>>>>> Are there any situations that the results of the XQuery could not be
<br>>>>> found
<br>>>>> in this way? I figured if the query returns the document it
<br>>>>> should be
<br>>>>> able
<br>>>>> to be deleted? The actual XQuery I'm using is pretty simple:
<br>>>>>
<br>>>>> (31 Jul 2006 21:55:10) DEBUG QueryPlan for XQuery
<br>>>>> [collection('dealers.dbxml')/als:dealer[@active = true()]]:
<br>>>>> <XQuery>
<br>>>>> <Navigation>
<br>>>>> <QueryPlanFunction result="collection"
<br>>>>> container="dealers.dbxml">
<br>>>>> <OQPlan>n(V(node-attribute-equality-
<br>>>>> boolean,@active,=,'true'),P(node-element-presence-
<br>>>>> none,=,dealer:<a href="http://autoleadservice.com/xml/als" target="_top" rel="nofollow">http://autoleadservice.com/xml/als</a>))</OQPlan>
<br>>>>> </QueryPlanFunction>
<br>>>>> <Step axis="child" uri="<a href="http://autoleadservice.com/xml/als" target="_top" rel="nofollow">http://autoleadservice.com/xml/als</a>"
<br>>>>> name="dealer" nodeType="element"/>
<br>>>>> <DbXmlFilter>
<br>>>>> <Navigation>
<br>>>>> <Step axis="attribute" name="active" nodeType="attribute"/>
<br>>>>> <DbXmlCompare name="equal">
<br>>>>> <Sequence>
<br>>>>> <AnyAtomicTypeConstructor value="true"
<br>>>>> typeuri="<a href="http://www.w3.org/2001/XMLSchema" target="_top" rel="nofollow">http://www.w3.org/2001/XMLSchema</a>"
<br>>>>> typename="boolean"/>
<br>>>>> </Sequence>
<br>>>>> </DbXmlCompare>
<br>>>>> </Navigation>
<br>>>>> </DbXmlFilter>
<br>>>>> </Navigation>
<br>>>>> </XQuery>
<br>>>>>
<br>>>>> Thanks for any insight...
<br>>>>>
<br>>>>> -- m@
<br>>>>>
<br>>>>>
<br>>>>>
<br>>>>> ------------------------------------------
<br>>>>> To remove yourself from this list, send an
<br>>>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5603782&i=0" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>>>
<br>>>>>
<br>>>>
<br>>>>
<br>>>>
<br>>>> ------------------------------------------
<br>>>> To remove yourself from this list, send an
<br>>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5603782&i=1" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>>
<br>>>>
<br>>>
<br>>>
<br>>>
<br>>> ------------------------------------------
<br>>> To remove yourself from this list, send an
<br>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5603782&i=2" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>
<br>>
<br>>
<br>> ------------------------------------------
<br>> To remove yourself from this list, send an
<br>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5603782&i=3" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>
<br>>
</div><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5603782&i=4" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5602496Re: DOCUMENT_NOT_FOUND when trying to update document after creating it within same transaction2006-08-01T14:17:42Z2006-08-01T14:17:42ZMatt Magoffin-4The code is not too large, no, I will send you a copy of that. As per your
<br>questions below:
<br><br>1) Might be possible in the general context of the application, but not
<br>within the scope of this transaction as a new XmlDocument is created for
<br>each object stored in the database.
<br><br>2) I have tried to find a situation where a transaction might not be used,
<br>and am fairly sure all operations occur within a transaction, including
<br>reads.
<br><br>3) I do have concurrent threads in the application, but this seems to
<br>occur even with a single thread operating.
<br><br>4) Yes, I have double-checked this, the error occurs after the
<br>getDocument() call, on the updateDocument() call.
<br><br>I will send you the code next.
<br><br>Thank you,
<br>m@
<br><div class='shrinkable-quote'><br>> Matt,
<br>>
<br>> Is the code too large to send? Missing pieces can provide
<br>> important clues. A few things to look for include:
<br>> 1. Avoid re-using XmlDocument objects without doing
<br>> a delete() on the previous use. This should work, but
<br>> it's safer to explicitly delete.
<br>> 2. Make sure you are using the right transactions on all operations,
<br>> including the reads.
<br>> 3. Do you have concurrent threads performing operations?
<br>> 4. Are you sure you are getting DOCUMENT_NOT_FOUND on
<br>> the updateDocument() call, and not the getDocument() call?
<br>>
<br>> Check these things and see if anything changes.
<br>>
<br>> Regards,
<br>>
<br>> George
<br>>
<br>> On Aug 1, 2006, at 4:39 AM, Matt Magoffin wrote:
<br>>
<br>>> I've been doing much testing on this, including completely deleting
<br>>> my DB
<br>>> environment and starting over from scratch. This pattern of
<br>>> updating works
<br>>> for some updates for some time, then seems to fail. Looking through
<br>>> my own
<br>>> debug logs, I see in a given transaction many hundreds of successful
<br>>> getDocument()/updateDocument() pairs, and then suddenly the
<br>>> DOCUMENT_NOT_FOUND exception will occur, seemingly at random.
<br>>>
<br>>> I think I've run out of ideas at this point.
<br>>>
<br>>> -- m@
<br>>>
<br>>>> Actually, this happens even if updating a document committed
<br>>>> already from
<br>>>> a previous transaction, too.
<br>>>>
<br>>>> Is the code pattern below improper in some way for updating an
<br>>>> existing
<br>>>> document during a transaction, i.e.
<br>>>>
<br>>>> String docId = ...;
<br>>>> String xml = ...;
<br>>>> XmlDocument xmlDoc = xmlContainer.getDocument(holder.getTransaction
<br>>>> (),
<br>>>> docId);
<br>>>> xmlDoc.setContent(xml);
<br>>>> xmlContainer.updateDocument(holder.getTransaction(), xmlDoc, null);
<br>>>>
<br>>>>> Hello, I ran into an exception
<br>>>>>
<br>>>>> errcode = DOCUMENT_NOT_FOUND
<br>>>>>
<br>>>>> during a transaction where previously in the transaction I have
<br>>>>> created
<br>>>>> an
<br>>>>> XmlDocument and stored it, and then later happen to try to update
<br>>>>> that
<br>>>>> document within the same transaction. The logic works as shown
<br>>>>> below.
<br>>>>> Basically within one transaction an XmlDocument is created as new
<br>>>>> with a
<br>>>>> xmlManager.putDocument() call, and then later in the transaction the
<br>>>>> same
<br>>>>> document is attempted to be changed with an
<br>>>>> xmlManager.getDocument()/xmlManager.updateDocument() pair of
<br>>>>> calls. Is
<br>>>>> this allowed?
<br>>>>>
<br>>>>> -----
<br>>>>> XmlDocument xmlDoc = null;
<br>>>>> try {
<br>>>>> xmlDoc = xmlContainer.getDocument(holder.getTransaction(), docId);
<br>>>>> } catch ( XmlException e ) {
<br>>>>> if ( e.getErrorCode() == XmlException.DOCUMENT_NOT_FOUND ) {
<br>>>>> xmlDoc = xmlManager.createDocument();
<br>>>>> } else {
<br>>>>> throw e;
<br>>>>> }
<br>>>>> }
<br>>>>>
<br>>>>> ...
<br>>>>>
<br>>>>> if ( StringUtils.hasLength(xmlDoc.getName()) ) {
<br>>>>> xmlContainer.updateDocument(holder.getTransaction(), xmlDoc,
<br>>>>> updateContext);
<br>>>>> } else {
<br>>>>> xmlDoc.setName(docId);
<br>>>>> xmlContainer.putDocument(holder.getTransaction(), xmlDoc,
<br>>>>> updateContext,
<br>>>>> docConfig);
<br>>>>> }
<br>>>>> -----
<br>>>>>
<br>>>>> So the first time around the XmlDocument would have been
<br>>>>>
<br>>>>>
<br>>>>> ------------------------------------------
<br>>>>> To remove yourself from this list, send an
<br>>>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5602496&i=0" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>>>
<br>>>>>
<br>>>>
<br>>>>
<br>>>>
<br>>>> ------------------------------------------
<br>>>> To remove yourself from this list, send an
<br>>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5602496&i=1" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>>
<br>>>>
<br>>>
<br>>>
<br>>>
<br>>> ------------------------------------------
<br>>> To remove yourself from this list, send an
<br>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5602496&i=2" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>
<br>>
<br>>
<br>>
<br>> ------------------------------------------
<br>> To remove yourself from this list, send an
<br>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5602496&i=3" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>
<br>>
</div><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5602496&i=4" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5599501Re: Unexpected Tomcat closure breaks containers?2006-08-01T11:08:14Z2006-08-01T11:08:14ZJohn Snelson-4Hi Antony,
<br><br>So long as you are using transactions, this situation is not a problem.
<br>Since DB XML does not run as a separate server, the containers can often
<br>be corrupted due to an application crash. In these situations, you will
<br>need to run the recovery process, which will replay the transaction
<br>logs. This webpage should help you out:
<br><br><a href="http://www.sleepycat.com/docs/ref/transapp/recovery.html" target="_top" rel="nofollow">http://www.sleepycat.com/docs/ref/transapp/recovery.html</a><br><br>Many people run recovery every time their applications start, since
<br>there is no penalty to running recovery when it is not needed.
<br><br>You should also be checkpointing, if you aren't already. This will
<br>reduce the amount of transaction logging that will need to be replayed:
<br><br><a href="http://www.sleepycat.com/docs/ref/transapp/checkpoint.html" target="_top" rel="nofollow">http://www.sleepycat.com/docs/ref/transapp/checkpoint.html</a><br><br>Hope that helps,
<br><br>John
<br><br>Antony Grinyer wrote:
<div class='shrinkable-quote'><br>> Hi,
<br>>
<br>> We are using BDB XML 2.2.13 on Windows XP with Java 1.5. We use Tomcat
<br>> for our web application and during start-up we open our containers using
<br>> a listener; we also close our containers using a listener when Tomcat
<br>> shuts down, we do not use the container 'auto-open' environment
<br>> parameter. During a recent demo, Tomcat crashed unexpectedly which
<br>> appeared to corrupt all of our containers when we started Tomcat again.
<br>> Therefore, my question is what is the solution to ensure containers are
<br>> kept uncorrupted given a Tomcat crash? Is there any 'auto-close on
<br>> crash' type parameters in BDB XML? Has anyone implemented a fail safe
<br>> scenario using Tomcat and BDB XML given a Tomcat crash?
<br>>
<br>> Thanks in advance for any help,
<br>>
<br>> Ant
<br>> ---
<br>> Senior Technical Consultant
<br>> CSW Group Ltd.
<br>>
<br>>
<br>>
<br>> ------------------------------------------
<br>> To remove yourself from this list, send an
<br>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5599501&i=0" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>
</div><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5599501&i=1" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5599370Re: Unexpected Tomcat closure breaks containers?2006-08-01T11:05:39Z2006-08-01T11:05:39ZMadhu SadagopanHi:<br><br>Enable EnvironmentConfig.setRunRecovery() <br>(Getting started guide on Java - Chapter 3)<br><br>or<br><br>Recover the XMLDB from the command prompt using 'db_recover' after any ungraceful shutdown.<br><br>Recovery is not possible if the logging subsytem is disabled as it generates the needed log files for recovery. Hope this helps.<br><br>bye,<br>Madhu<br><br><br><br><b><i>Antony Grinyer <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5599370&i=0" target="_top" rel="nofollow">antony.grinyer@...</a>></i></b> wrote:<blockquote class="replbq" style="border-left: 2px solid rgb(16, 16, 255); margin-left: 5px; padding-left: 5px;"> Hi,<br><br>We are using BDB XML 2.2.13 on Windows XP with Java 1.5. We use Tomcat<br>for our web application and during start-up we open our containers using<br>a listener; we also close our containers using a listener when Tomcat<br>shuts down, we do not use the container 'auto-open' environment<br>parameter. During a recent demo, Tomcat crashed unexpectedly which<br>appeared to corrupt all of our containers when we
started Tomcat again.<br>Therefore, my question is what is the solution to ensure containers are<br>kept uncorrupted given a Tomcat crash? Is there any 'auto-close on<br>crash' type parameters in BDB XML? Has anyone implemented a fail safe<br>scenario using Tomcat and BDB XML given a Tomcat crash?<br><br>Thanks in advance for any help,<br><br>Ant<br>---<br>Senior Technical Consultant<br>CSW Group Ltd.<br><br><br><br>------------------------------------------<br>To remove yourself from this list, send an<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5599370&i=1" target="_top" rel="nofollow">xml-unsubscribe@...</a><br><br></blockquote><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5598730Re: DB XML on FreeBSD2006-08-01T10:21:46Z2006-08-01T10:21:46ZMohamad Babaei<div>Thanks George,</div>
<div>it looks OK !</div>
<div>You're right,</div>
<div>i had installed gmake, but didn't use "-m gmake" with "buildall.sh" .</div>
<div>so i simply installed it like this :</div>
<div>sh buildall.sh -m gmake < options></div>
<div> </div>
<div>Best Regards,</div>
<div>-Mohamad<br><br> </div>
<div><span class="gmail_quote">On 8/1/06, <b class="gmail_sendername">George Feinberg</b> <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5598730&i=0" target="_top" rel="nofollow">george.feinberg@...</a>> wrote:</span>
<blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid"><br>> Thanks George,<br>> But i've already installed it ,<br>> In my previouse test i didn't say that the error i get, is in
<br>> "xerces" build time,<br><br>I believe that you are still picking up the wrong make.<br>Xerces definitely requires gnu make.<br>It can be helpful to install gnu make as "gmake" and<br>be explicit in
buildall.sh:<br> buildall.sh -m gmake <other options><br><br>George<br><div class='shrinkable-quote'><br>><br>> Best Regards,<br>> -Mohamad<br>><br>> On 8/1/06, George Feinberg <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5598730&i=1" target="_top" rel="nofollow">george.feinberg@...</a>> wrote:<br>> Mohamad,<br>><br>> You need to use gnu make to build BDB XML and its<br>> dependent libraries.<br>><br>> Regards,<br>><br>> George<br>><br>> On Aug 1, 2006, at 4:35 AM, Mohamad Babaei wrote:
<br>><br>> ><br>> ><br>> > On 7/31/06, Danny Brian <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5598730&i=2" target="_top" rel="nofollow">dan@...</a>> wrote: > Hi,<br>> > > Has anybody installed DB XML on FreeBSD ?<br>> > > please let me know about the installation proccess,
<br>> ><br>> > It's documented in the Unix build instructions.<br>> ><br>> > Be sure and use gcc 4.1 or later, and include them in the arguments<br>> > to buildall.sh.<br>> ><br>> > - Danny
<br>> ><br>> ><br>> > Thank you !<br>> > but there is some problems with it :<br>> > .<br>> > .<br>> > .<br>> > .<br>> > "Makefile", line 252: Need an operator
<br>> > "Makefile", line 253: Need an operator<br>> > "Makefile", line 254: Missing dependency operator<br>> > "Makefile", line 255: warning: duplicate script for target "ifeq"
<br>> > ignored<br>> > "Makefile", line 255: warning: duplicate script for target<br>> > "(INMEM,ICU)" ignored<br>> > "Makefile", line 256: Need an operator<br>> > "Makefile", line 257: Missing dependency operator
<br>> > "Makefile", line 258: warning: duplicate script for target "ifeq"<br>> > ignored<br>> > "Makefile", line 258: warning: duplicate script for target<br>> > "(INMEM,ICONV)" ignored
<br>> > "Makefile", line 259: Need an operator<br>> > make: fatal errors encountered -- cannot continue<br>> ><br>> > would you please give me more detailed instructions ?<br>> ></div>
> > Regards,<br>> > Mo<br>> ><br>><br>><br><br></blockquote></div><br>
<p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5597918Unexpected Tomcat closure breaks containers?2006-08-01T09:29:45Z2006-08-01T09:29:45ZAntony GrinyerHi,
<br><br>We are using BDB XML 2.2.13 on Windows XP with Java 1.5. We use Tomcat
<br>for our web application and during start-up we open our containers using
<br>a listener; we also close our containers using a listener when Tomcat
<br>shuts down, we do not use the container 'auto-open' environment
<br>parameter. During a recent demo, Tomcat crashed unexpectedly which
<br>appeared to corrupt all of our containers when we started Tomcat again.
<br>Therefore, my question is what is the solution to ensure containers are
<br>kept uncorrupted given a Tomcat crash? Is there any 'auto-close on
<br>crash' type parameters in BDB XML? Has anyone implemented a fail safe
<br>scenario using Tomcat and BDB XML given a Tomcat crash?
<br><br>Thanks in advance for any help,
<br><br>Ant
<br>---
<br>Senior Technical Consultant
<br>CSW Group Ltd.
<br><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5597918&i=0" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5597492Re: DOCUMENT_NOT_FOUND when trying to update document after creating it within same transaction2006-08-01T09:26:54Z2006-08-01T09:26:54ZGeorge Feinberg-3Matt,
<br><br>Is the code too large to send? Missing pieces can provide
<br>important clues. A few things to look for include:
<br>1. Avoid re-using XmlDocument objects without doing
<br>a delete() on the previous use. This should work, but
<br>it's safer to explicitly delete.
<br>2. Make sure you are using the right transactions on all operations,
<br>including the reads.
<br>3. Do you have concurrent threads performing operations?
<br>4. Are you sure you are getting DOCUMENT_NOT_FOUND on
<br>the updateDocument() call, and not the getDocument() call?
<br><br>Check these things and see if anything changes.
<br><br>Regards,
<br><br>George
<br><br>On Aug 1, 2006, at 4:39 AM, Matt Magoffin wrote:
<br><div class='shrinkable-quote'><br>> I've been doing much testing on this, including completely deleting
<br>> my DB
<br>> environment and starting over from scratch. This pattern of
<br>> updating works
<br>> for some updates for some time, then seems to fail. Looking through
<br>> my own
<br>> debug logs, I see in a given transaction many hundreds of successful
<br>> getDocument()/updateDocument() pairs, and then suddenly the
<br>> DOCUMENT_NOT_FOUND exception will occur, seemingly at random.
<br>>
<br>> I think I've run out of ideas at this point.
<br>>
<br>> -- m@
<br>>
<br>>> Actually, this happens even if updating a document committed
<br>>> already from
<br>>> a previous transaction, too.
<br>>>
<br>>> Is the code pattern below improper in some way for updating an
<br>>> existing
<br>>> document during a transaction, i.e.
<br>>>
<br>>> String docId = ...;
<br>>> String xml = ...;
<br>>> XmlDocument xmlDoc = xmlContainer.getDocument(holder.getTransaction
<br>>> (),
<br>>> docId);
<br>>> xmlDoc.setContent(xml);
<br>>> xmlContainer.updateDocument(holder.getTransaction(), xmlDoc, null);
<br>>>
<br>>>> Hello, I ran into an exception
<br>>>>
<br>>>> errcode = DOCUMENT_NOT_FOUND
<br>>>>
<br>>>> during a transaction where previously in the transaction I have
<br>>>> created
<br>>>> an
<br>>>> XmlDocument and stored it, and then later happen to try to update
<br>>>> that
<br>>>> document within the same transaction. The logic works as shown
<br>>>> below.
<br>>>> Basically within one transaction an XmlDocument is created as new
<br>>>> with a
<br>>>> xmlManager.putDocument() call, and then later in the transaction the
<br>>>> same
<br>>>> document is attempted to be changed with an
<br>>>> xmlManager.getDocument()/xmlManager.updateDocument() pair of
<br>>>> calls. Is
<br>>>> this allowed?
<br>>>>
<br>>>> -----
<br>>>> XmlDocument xmlDoc = null;
<br>>>> try {
<br>>>> xmlDoc = xmlContainer.getDocument(holder.getTransaction(), docId);
<br>>>> } catch ( XmlException e ) {
<br>>>> if ( e.getErrorCode() == XmlException.DOCUMENT_NOT_FOUND ) {
<br>>>> xmlDoc = xmlManager.createDocument();
<br>>>> } else {
<br>>>> throw e;
<br>>>> }
<br>>>> }
<br>>>>
<br>>>> ...
<br>>>>
<br>>>> if ( StringUtils.hasLength(xmlDoc.getName()) ) {
<br>>>> xmlContainer.updateDocument(holder.getTransaction(), xmlDoc,
<br>>>> updateContext);
<br>>>> } else {
<br>>>> xmlDoc.setName(docId);
<br>>>> xmlContainer.putDocument(holder.getTransaction(), xmlDoc,
<br>>>> updateContext,
<br>>>> docConfig);
<br>>>> }
<br>>>> -----
<br>>>>
<br>>>> So the first time around the XmlDocument would have been
<br>>>>
<br>>>>
<br>>>> ------------------------------------------
<br>>>> To remove yourself from this list, send an
<br>>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5597492&i=0" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>>
<br>>>>
<br>>>
<br>>>
<br>>>
<br>>> ------------------------------------------
<br>>> To remove yourself from this list, send an
<br>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5597492&i=1" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>
<br>>>
<br>>
<br>>
<br>>
<br>> ------------------------------------------
<br>> To remove yourself from this list, send an
<br>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5597492&i=2" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>
</div><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5597492&i=3" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5596724Reminder: Berkeley DB XML mailing list transition2006-08-01T09:01:47Z2006-08-01T09:01:47ZGeorge Feinberg-3A reminder. Traffic on this mailing list should move to the
<br>Berkeley DB XML forum:
<br> <a href="http://forums.oracle.com/forums/forum.jspa?forumID=274" target="_top" rel="nofollow">http://forums.oracle.com/forums/forum.jspa?forumID=274</a><br><br>All of the Sleepycat Software Berkeley DB support channels will move
<br>to Oracle support channels
<br>during the month of August.
<br><br>This includes the Berkeley DB Java Edition, Berkeley DB XML and
<br>Berkeley DB High Availability product mailing lists.
<br><br>The mailing lists are being moved to the Oracle Technical Network
<br>forums:
<br> <a href="http://forums.oracle.com/forums/category.jspa?categoryID=18" target="_top" rel="nofollow">http://forums.oracle.com/forums/category.jspa?categoryID=18</a><br><br>There are four technical forums at this site: one each for Berkeley
<br>DB, Berkeley DB High Availability, Berkeley DB XML and Berkeley DB
<br>Java Edition. Please select the appropriate forum and post your
<br>question there.
<br><br>The Berkeley DB support and engineering teams are already monitoring
<br>these forums, and you can expect prompt and helpful answers.
<br><br>After an appropriate transition period, the Berkeley DB mailing lists
<br>will be turned off.
<br><br>While it will no longer be possible to send email directly to a mailing
<br>list, the use of an OTN forum has a couple of features that you may
<br>appreciate:
<br>1. The forum is searchable
<br>2. An RSS feed is available
<br>3. You can "watch" a forum, getting email notifications of new
<br>postings,
<br>with the content of the post.
<br>4. You can also watch specific threads or users.
<br>5. You can read, and search the forum as "guest" but you need to
<br>register
<br>and sign in to post.
<br><br>Regards,
<br><br>George Feinberg
<br><br><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5596724&i=0" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---Announce-f731.html" embed="fixTarget[731]" target="_top" >Berkeley DB Xml - Announce</a></p>tag:old.nabble.com,2006:post-5596375Reminder: Berkeley DB XML mailing list transition2006-08-01T09:01:47Z2006-08-01T09:01:47ZGeorge Feinberg-3A reminder. Traffic on this mailing list should move to the
<br>Berkeley DB XML forum:
<br> <a href="http://forums.oracle.com/forums/forum.jspa?forumID=274" target="_top" rel="nofollow">http://forums.oracle.com/forums/forum.jspa?forumID=274</a><br><br>All of the Sleepycat Software Berkeley DB support channels will move
<br>to Oracle support channels
<br>during the month of August.
<br><br>This includes the Berkeley DB Java Edition, Berkeley DB XML and
<br>Berkeley DB High Availability product mailing lists.
<br><br>The mailing lists are being moved to the Oracle Technical Network
<br>forums:
<br> <a href="http://forums.oracle.com/forums/category.jspa?categoryID=18" target="_top" rel="nofollow">http://forums.oracle.com/forums/category.jspa?categoryID=18</a><br><br>There are four technical forums at this site: one each for Berkeley
<br>DB, Berkeley DB High Availability, Berkeley DB XML and Berkeley DB
<br>Java Edition. Please select the appropriate forum and post your
<br>question there.
<br><br>The Berkeley DB support and engineering teams are already monitoring
<br>these forums, and you can expect prompt and helpful answers.
<br><br>After an appropriate transition period, the Berkeley DB mailing lists
<br>will be turned off.
<br><br>While it will no longer be possible to send email directly to a mailing
<br>list, the use of an OTN forum has a couple of features that you may
<br>appreciate:
<br>1. The forum is searchable
<br>2. An RSS feed is available
<br>3. You can "watch" a forum, getting email notifications of new
<br>postings,
<br>with the content of the post.
<br>4. You can also watch specific threads or users.
<br>5. You can read, and search the forum as "guest" but you need to
<br>register
<br>and sign in to post.
<br><br>Regards,
<br><br>George Feinberg
<br><br><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5596375&i=0" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5596375&i=1" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5595182Re: DB XML on FreeBSD2006-08-01T08:00:12Z2006-08-01T08:00:12ZGeorge Feinberg-3<br>> Thanks George,
<br>> But i've already installed it ,
<br>> In my previouse test i didn't say that the error i get, is in
<br>> "xerces" build time,
<br><br>I believe that you are still picking up the wrong make.
<br>Xerces definitely requires gnu make.
<br>It can be helpful to install gnu make as "gmake" and
<br>be explicit in buildall.sh:
<br> buildall.sh -m gmake <other options>
<br><br>George
<br><div class='shrinkable-quote'><br>>
<br>> Best Regards,
<br>> -Mohamad
<br>>
<br>> On 8/1/06, George Feinberg <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5595182&i=0" target="_top" rel="nofollow">george.feinberg@...</a>> wrote:
<br>> Mohamad,
<br>>
<br>> You need to use gnu make to build BDB XML and its
<br>> dependent libraries.
<br>>
<br>> Regards,
<br>>
<br>> George
<br>>
<br>> On Aug 1, 2006, at 4:35 AM, Mohamad Babaei wrote:
<br>>
<br>> >
<br>> >
<br>> > On 7/31/06, Danny Brian <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5595182&i=1" target="_top" rel="nofollow">dan@...</a>> wrote: > Hi,
<br>> > > Has anybody installed DB XML on FreeBSD ?
<br>> > > please let me know about the installation proccess,
<br>> >
<br>> > It's documented in the Unix build instructions.
<br>> >
<br>> > Be sure and use gcc 4.1 or later, and include them in the arguments
<br>> > to buildall.sh.
<br>> >
<br>> > - Danny
<br>> >
<br>> >
<br>> > Thank you !
<br>> > but there is some problems with it :
<br>> > .
<br>> > .
<br>> > .
<br>> > .
<br>> > "Makefile", line 252: Need an operator
<br>> > "Makefile", line 253: Need an operator
<br>> > "Makefile", line 254: Missing dependency operator
<br>> > "Makefile", line 255: warning: duplicate script for target "ifeq"
<br>> > ignored
<br>> > "Makefile", line 255: warning: duplicate script for target
<br>> > "(INMEM,ICU)" ignored
<br>> > "Makefile", line 256: Need an operator
<br>> > "Makefile", line 257: Missing dependency operator
<br>> > "Makefile", line 258: warning: duplicate script for target "ifeq"
<br>> > ignored
<br>> > "Makefile", line 258: warning: duplicate script for target
<br>> > "(INMEM,ICONV)" ignored
<br>> > "Makefile", line 259: Need an operator
<br>> > make: fatal errors encountered -- cannot continue
<br>> >
<br>> > would you please give me more detailed instructions ?
<br>> >
<br>> > Regards,
<br>> > Mo
<br>> >
<br>>
<br>>
</div><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5595182&i=2" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5594975Re: DB XML on FreeBSD2006-08-01T07:50:46Z2006-08-01T07:50:46ZMohamad Babaei<div>Thanks George,</div>
<div>But i've already installed it ,</div>
<div>In my previouse test i didn't say that the error i get, is in "xerces" build time,</div>
<div> </div>
<div>Best Regards,</div>
<div>-Mohamad<br> </div>
<div><span class="gmail_quote">On 8/1/06, <b class="gmail_sendername">George Feinberg</b> <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5594975&i=0" target="_top" rel="nofollow">george.feinberg@...</a>> wrote:</span>
<blockquote class="gmail_quote" style="PADDING-LEFT: 1ex; MARGIN: 0px 0px 0px 0.8ex; BORDER-LEFT: #ccc 1px solid">Mohamad,<br><br>You need to use gnu make to build BDB XML and its<br>dependent libraries.<br><br>Regards,<br>
<br>George<br><br>On Aug 1, 2006, at 4:35 AM, Mohamad Babaei wrote:<br><div class='shrinkable-quote'><br>><br>><br>> On 7/31/06, Danny Brian <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5594975&i=1" target="_top" rel="nofollow">dan@...</a>> wrote: > Hi,<br>> > Has anybody installed DB XML on FreeBSD ?
<br>> > please let me know about the installation proccess,<br>><br>> It's documented in the Unix build instructions.<br>><br>> Be sure and use gcc 4.1 or later, and include them in the arguments<br>> to
buildall.sh.<br>><br>> - Danny<br>><br>><br>> Thank you !<br>> but there is some problems with it :<br>> .<br>> .<br>> .<br>> .<br>> "Makefile", line 252: Need an operator<br>> "Makefile", line 253: Need an operator
<br>> "Makefile", line 254: Missing dependency operator<br>> "Makefile", line 255: warning: duplicate script for target "ifeq"<br>> ignored<br>> "Makefile", line 255: warning: duplicate script for target
<br>> "(INMEM,ICU)" ignored<br>> "Makefile", line 256: Need an operator<br>> "Makefile", line 257: Missing dependency operator<br>> "Makefile", line 258: warning: duplicate script for target "ifeq"
<br>> ignored<br>> "Makefile", line 258: warning: duplicate script for target<br>> "(INMEM,ICONV)" ignored<br>> "Makefile", line 259: Need an operator<br>> make: fatal errors encountered -- cannot continue
<br>><br>> would you please give me more detailed instructions ?<br>><br>> Regards,<br>> Mo<br>></div><br></blockquote></div><br>
<p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5594721Re: DB XML on FreeBSD2006-08-01T07:34:25Z2006-08-01T07:34:25ZGeorge Feinberg-3Mohamad,
<br><br>You need to use gnu make to build BDB XML and its
<br>dependent libraries.
<br><br>Regards,
<br><br>George
<br><br>On Aug 1, 2006, at 4:35 AM, Mohamad Babaei wrote:
<br><div class='shrinkable-quote'><br>>
<br>>
<br>> On 7/31/06, Danny Brian <<a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5594721&i=0" target="_top" rel="nofollow">dan@...</a>> wrote: > Hi,
<br>> > Has anybody installed DB XML on FreeBSD ?
<br>> > please let me know about the installation proccess,
<br>>
<br>> It's documented in the Unix build instructions.
<br>>
<br>> Be sure and use gcc 4.1 or later, and include them in the arguments
<br>> to buildall.sh.
<br>>
<br>> - Danny
<br>>
<br>>
<br>> Thank you !
<br>> but there is some problems with it :
<br>> .
<br>> .
<br>> .
<br>> .
<br>> "Makefile", line 252: Need an operator
<br>> "Makefile", line 253: Need an operator
<br>> "Makefile", line 254: Missing dependency operator
<br>> "Makefile", line 255: warning: duplicate script for target "ifeq"
<br>> ignored
<br>> "Makefile", line 255: warning: duplicate script for target
<br>> "(INMEM,ICU)" ignored
<br>> "Makefile", line 256: Need an operator
<br>> "Makefile", line 257: Missing dependency operator
<br>> "Makefile", line 258: warning: duplicate script for target "ifeq"
<br>> ignored
<br>> "Makefile", line 258: warning: duplicate script for target
<br>> "(INMEM,ICONV)" ignored
<br>> "Makefile", line 259: Need an operator
<br>> make: fatal errors encountered -- cannot continue
<br>>
<br>> would you please give me more detailed instructions ?
<br>>
<br>> Regards,
<br>> Mo
<br>>
</div><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5594721&i=1" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5594454Re: DOCUMENT_NOT_FOUND when try to delete while iterating over XQuery result set2006-08-01T07:25:46Z2006-08-01T07:25:46ZGeorge Feinberg-3Matt,
<br><br>The most obvious explanation is that you are getting more
<br>than one result for that document, and removing it
<br>based on the first result. Then, when you process
<br>your second result, the document's gone.
<br><br>If this is possible, just catch, and ignore this exception.
<br><br>You can change your deleteDocument() call to one
<br>that just prints out the document name to see if
<br>this is happening.
<br><br>George
<br><br>On Jul 31, 2006, at 10:59 PM, Matt Magoffin wrote:
<br><div class='shrinkable-quote'><br>> Sorry, actually I spoke too soon. I still see the below exception,
<br>> even
<br>> with DB 4.4.
<br>>
<br>> Any insight would be appreciated!
<br>>
<br>> -- m@
<br>>
<br>>> I recompiled BDB XML against DB 4.4.20, and the problems seem to
<br>>> have gone
<br>>> away. I had been using the DB 4.3 that came with XML 2.2.13
<br>>> previously.
<br>>>
<br>>> -- m@
<br>>>
<br>>>> Hello,
<br>>>>
<br>>>> I've been having a problem appear every so often that I can't
<br>>>> seem to
<br>>>> track down. It seems to happen more frequently after loading sets of
<br>>>> data,
<br>>>> but I cannot be sure exactly of the cause.
<br>>>>
<br>>>> The symptom is that I execute an XQuery and then try to delete
<br>>>> all the
<br>>>> documents resulting from the XQuery, all within a transaction. If
<br>>>> I step
<br>>>> through the code in a debugger, I see BDB XML returning the
<br>>>> XmlDocument
<br>>>> instances, but then throws the below exception:
<br>>>>
<br>>>> -----
<br>>>> Error: Document not found: 4991476, errcode = DOCUMENT_NOT_FOUND
<br>>>> at
<br>>>> com.sleepycat.dbxml.dbxml_javaJNI.XmlContainer_deleteDocument__SWIG_
<br>>>> 2(Native
<br>>>> Method)
<br>>>> at
<br>>>> com.sleepycat.dbxml.XmlContainer.deleteDocument(XmlContainer.java:
<br>>>> 367)
<br>>>> -----
<br>>>>
<br>>>> The general flow of the program is this:
<br>>>>
<br>>>> -----
<br>>>> XmlQueryContext context = xmlManager.createQueryContext();
<br>>>> XmlQueryExpression qe = createQueryExpression(criteria, context,
<br>>>> XmlQueryContext.LiveValues, XmlQueryContext.Lazy);
<br>>>> XmlResults results = qe.execute(holder.getTransaction(), context);
<br>>>>
<br>>>> XmlUpdateContext updateContext = xmlManager.createUpdateContext();
<br>>>> XmlDocument docToDelete = xmlManager.createDocument();
<br>>>> while ( results.next(docToDelete) ) {
<br>>>> xmlContainer.deleteDocument(holder.getTransaction(),
<br>>>> docToDelete, updateContext);
<br>>>> }
<br>>>> -----
<br>>>>
<br>>>> I've tried experimenting with XmlQueryContext.Eager, but that did
<br>>>> not
<br>>>> seem
<br>>>> to make any difference.
<br>>>>
<br>>>> Are there any situations that the results of the XQuery could not be
<br>>>> found
<br>>>> in this way? I figured if the query returns the document it
<br>>>> should be
<br>>>> able
<br>>>> to be deleted? The actual XQuery I'm using is pretty simple:
<br>>>>
<br>>>> (31 Jul 2006 21:55:10) DEBUG QueryPlan for XQuery
<br>>>> [collection('dealers.dbxml')/als:dealer[@active = true()]]:
<br>>>> <XQuery>
<br>>>> <Navigation>
<br>>>> <QueryPlanFunction result="collection"
<br>>>> container="dealers.dbxml">
<br>>>> <OQPlan>n(V(node-attribute-equality-
<br>>>> boolean,@active,=,'true'),P(node-element-presence-
<br>>>> none,=,dealer:<a href="http://autoleadservice.com/xml/als" target="_top" rel="nofollow">http://autoleadservice.com/xml/als</a>))</OQPlan>
<br>>>> </QueryPlanFunction>
<br>>>> <Step axis="child" uri="<a href="http://autoleadservice.com/xml/als" target="_top" rel="nofollow">http://autoleadservice.com/xml/als</a>"
<br>>>> name="dealer" nodeType="element"/>
<br>>>> <DbXmlFilter>
<br>>>> <Navigation>
<br>>>> <Step axis="attribute" name="active" nodeType="attribute"/>
<br>>>> <DbXmlCompare name="equal">
<br>>>> <Sequence>
<br>>>> <AnyAtomicTypeConstructor value="true"
<br>>>> typeuri="<a href="http://www.w3.org/2001/XMLSchema" target="_top" rel="nofollow">http://www.w3.org/2001/XMLSchema</a>"
<br>>>> typename="boolean"/>
<br>>>> </Sequence>
<br>>>> </DbXmlCompare>
<br>>>> </Navigation>
<br>>>> </DbXmlFilter>
<br>>>> </Navigation>
<br>>>> </XQuery>
<br>>>>
<br>>>> Thanks for any insight...
<br>>>>
<br>>>> -- m@
<br>>>>
<br>>>>
<br>>>>
<br>>>> ------------------------------------------
<br>>>> To remove yourself from this list, send an
<br>>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5594454&i=0" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>>
<br>>>>
<br>>>
<br>>>
<br>>>
<br>>> ------------------------------------------
<br>>> To remove yourself from this list, send an
<br>>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5594454&i=1" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>>
<br>>>
<br>>
<br>>
<br>>
<br>> ------------------------------------------
<br>> To remove yourself from this list, send an
<br>> email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5594454&i=2" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br>>
</div><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5594454&i=3" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>tag:old.nabble.com,2006:post-5594270Re: Xml DB2006-08-01T07:13:18Z2006-08-01T07:13:18ZGeorge Feinberg-3Marc,
<br><br>There are unfortunately no examples of Berkeley DB replication available
<br>in Java or for BDB XML. There are examples
<br>in C in the Berkeley DB tree. See db-4.3.29/examples_c/ex_repquote
<br>and ex_rep.
<br><br>There are several users who have done replication in BDB XML.
<br>Maybe one of them will chime in.
<br><br>Regards,
<br><br>George
<br><br><div class='shrinkable-quote'><br>> I'm in the process of evaluating Berkeley DBXML for an embedded xml
<br>> data store for the bank i work for (Société Générale in Paris).
<br>>
<br>> The goal is to accelerate report production from massive xml source
<br>> (9gigs of XML growing fast).
<br>>
<br>> To make this solution viable, i need to implement replication of
<br>> dbxml database over the network, and i can't find any example how
<br>> to implement replication in java. Application run on 2 J2EE servers.
<br>>
<br>> Can you help me getting started? I would really appreciate.
<br>> I attached a java file of were i'm at actually if it can help
<br>> getting me started.
<br>>
<br>> Thanks in advance.
<br>>
<br>> Marc Deschamps
<br>> Loan Studio Project Manager
<br>> <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5594270&i=0" target="_top" rel="nofollow">marc.deschamps@...</a>
<br>>
<br>> package core;
<br>>
<br>> import com.sgcib.lso.dbxml.MyXMLDBEnv;
<br>> import com.sleepycat.dbxml.XmlContainer;
<br>> import com.sleepycat.dbxml.XmlDocument;
<br>> import com.sleepycat.dbxml.XmlDocumentConfig;
<br>> import com.sleepycat.dbxml.XmlManager;
<br>> import com.sleepycat.dbxml.XmlUpdateContext;
<br>> import java.util.LinkedList;
<br>> import org.jgroups.Address;
<br>> import org.jgroups.Channel;
<br>> import org.jgroups.JChannel;
<br>> import org.jgroups.MembershipListener;
<br>> import org.jgroups.Message;
<br>> import org.jgroups.MessageListener;
<br>> import org.jgroups.View;
<br>> import org.jgroups.blocks.PullPushAdapter;
<br>> import org.jgroups.util.Util;
<br>>
<br>> /**
<br>> *
<br>> * @author Marc
<br>> */
<br>> public class Sender implements Runnable
<br>> {
<br>>
<br>> public Sender()
<br>> {
<br>> super();
<br>> }
<br>>
<br>> public static void main(String[] args)
<br>> {
<br>> try
<br>> {
<br>> Sender main = new Sender();
<br>> main.run();
<br>> }
<br>> catch (Exception e)
<br>> {
<br>> System.err.println(e.getMessage());
<br>> e.printStackTrace(System.err);
<br>> }
<br>> }
<br>>
<br>> public void run()
<br>> {
<br>> try
<br>> {
<br>> MyXMLDBEnv env = new MyXMLDBEnv("c:/www/data");
<br>> XmlManager mgr = env.getManager();
<br>> XmlContainer container = null;
<br>> try
<br>> {
<br>> container = mgr.openContainer("test.dbxml");
<br>> }
<br>> catch (Exception e)
<br>> {
<br>> container = mgr.createContainer("test.dbxml");
<br>> }
<br>> XmlDocumentConfig docConfig =
<br>> XmlDocumentConfig.DEFAULT.setGenerateName(true);
<br>> XmlUpdateContext uCtx = mgr.createUpdateContext();
<br>> for(int i = 0; i < 100000; i++)
<br>> {
<br>> XmlDocument doc = mgr.createDocument();
<br>> doc.setContent("<test>test" + i + "</test>");
<br>> container.putDocument(doc, uCtx, docConfig);
<br>> }
<br>> System.out.println("sended 100000 documents");
<br>> while(true)
<br>> {
<br>> Thread.sleep(1000);
<br>> }
<br>> }
<br>> catch (Exception e)
<br>> {
<br>> System.err.println(e.getMessage());
<br>> e.printStackTrace(System.err);
<br>> }
<br>> return;
<br>> }
<br>> }
<br>> package core;
<br>>
<br>> import com.sgcib.lso.dbxml.MyXMLDBEnv;
<br>> import com.sleepycat.dbxml.XmlContainer;
<br>> import com.sleepycat.dbxml.XmlDocument;
<br>> import com.sleepycat.dbxml.XmlDocumentConfig;
<br>> import com.sleepycat.dbxml.XmlManager;
<br>> import com.sleepycat.dbxml.XmlUpdateContext;
<br>> import java.util.LinkedList;
<br>> import org.jgroups.Address;
<br>> import org.jgroups.Channel;
<br>> import org.jgroups.JChannel;
<br>> import org.jgroups.MembershipListener;
<br>> import org.jgroups.Message;
<br>> import org.jgroups.MessageListener;
<br>> import org.jgroups.View;
<br>> import org.jgroups.blocks.PullPushAdapter;
<br>> import org.jgroups.util.Util;
<br>>
<br>> /**
<br>> *
<br>> * @author Marc
<br>> */
<br>> public class Receiver implements Runnable
<br>> {
<br>>
<br>> public Receiver()
<br>> {
<br>> super();
<br>> }
<br>>
<br>> public static void main(String[] args)
<br>> {
<br>> try
<br>> {
<br>> Receiver main = new Receiver();
<br>> main.run();
<br>> }
<br>> catch (Exception e)
<br>> {
<br>> System.err.println(e.getMessage());
<br>> e.printStackTrace(System.err);
<br>> }
<br>> }
<br>>
<br>> public void run()
<br>> {
<br>> try
<br>> {
<br>> MyXMLDBEnv env = new MyXMLDBEnv("c:/www/data_replicated");
<br>> XmlManager mgr = env.getManager();
<br>> XmlContainer container = null;
<br>> try
<br>> {
<br>> container = mgr.openContainer("test.dbxml");
<br>> }
<br>> catch (Exception e)
<br>> {
<br>> container = mgr.createContainer("test.dbxml");
<br>> }
<br>> while(true)
<br>> {
<br>> Thread.sleep(1000);
<br>> }
<br>> }
<br>> catch (Exception e)
<br>> {
<br>> System.err.println(e.getMessage());
<br>> e.printStackTrace(System.err);
<br>> }
<br>> return;
<br>> }
<br>> }
<br>> package com.sgcib.lso.dbxml;
<br>>
<br>> import com.sleepycat.db.DatabaseEntry;
<br>> import com.sleepycat.db.LogSequenceNumber;
<br>> import com.sleepycat.db.ReplicationTransport;
<br>> import com.sleepycat.db.internal.DbConstants;
<br>> import com.sleepycat.dbxml.XmlManagerConfig;
<br>> import java.io.File;
<br>>
<br>> import com.sleepycat.db.DatabaseException;
<br>> import com.sleepycat.db.Environment;
<br>> import com.sleepycat.db.EnvironmentConfig;
<br>> import com.sleepycat.dbxml.XmlManager;
<br>> import java.io.Serializable;
<br>> import java.util.LinkedList;
<br>> import org.jgroups.Address;
<br>> import org.jgroups.Channel;
<br>> import org.jgroups.JChannel;
<br>> import org.jgroups.MembershipListener;
<br>> import org.jgroups.Message;
<br>> import org.jgroups.MessageListener;
<br>> import org.jgroups.View;
<br>> import org.jgroups.blocks.LockingException;
<br>> import org.jgroups.blocks.PullPushAdapter;
<br>> import org.jgroups.blocks.ReplicationManager;
<br>> import org.jgroups.blocks.ReplicationReceiver;
<br>> import org.jgroups.blocks.UpdateException;
<br>> import org.jgroups.blocks.Xid;
<br>> import org.jgroups.util.Util;
<br>>
<br>> //Class used to open and close a Berkeley DB environment
<br>> public class MyXMLDBEnv implements ReplicationTransport,
<br>> MessageListener, MembershipListener
<br>> {
<br>> // database
<br>> private Environment dbEnv = null;
<br>> private XmlManager mgr = null;
<br>> private boolean dbEnvIsOpen = false;
<br>> private File home = null;
<br>> private LinkedList history = new LinkedList();
<br>>
<br>> //Replication
<br>> private Integer id;
<br>> private Integer masterId;
<br>>
<br>> private Channel channel;
<br>> private PullPushAdapter adapter;
<br>> private final String group_name = "BerkeleyDBXMLGroup";
<br>> private final String props="UDP
<br>> (mcast_addr=228.10.10.10;mcast_port=45566;bind_addr=127.0.0.1):"
<br>> + "PING:"
<br>> + "FD:"
<br>> + "STABLE:"
<br>> + "NAKACK:"
<br>> + "UNICAST:"
<br>> + "FRAG:"
<br>> + "FLUSH:"
<br>> + "GMS:"
<br>> + "VIEW_ENFORCER:"
<br>> + "STATE_TRANSFER:"
<br>> + "QUEUE";
<br>>
<br>> public MyXMLDBEnv(String path2DbEnv)
<br>> {
<br>> try
<br>> {
<br>> home = new File(path2DbEnv);
<br>> if (!home.isDirectory())
<br>> {
<br>> throw new Exception(home.getPath() + " does not
<br>> exist or is not a directory.");
<br>> }
<br>>
<br>> channel=new JChannel(props);
<br>> channel.setOpt(Channel.AUTO_RECONNECT, Boolean.TRUE);
<br>> channel.setOpt(Channel.AUTO_GETSTATE, Boolean.TRUE);
<br>> System.out.println("Connecting to " + group_name);
<br>> channel.connect(group_name);
<br>>
<br>> adapter=new PullPushAdapter(channel, this, this);
<br>> channel.getState(null, 5000);
<br>> id = new Integer((int) channel.getView().getVid().getId
<br>> ());
<br>>
<br>> EnvironmentConfig config = EnvironmentConfig.DEFAULT;
<br>> config.setCacheSize(50 * 1024 * 1024);
<br>> config.setAllowCreate(true);
<br>> config.setInitializeCache(true);
<br>> config.setTransactional(true);
<br>> config.setInitializeLocking(true);
<br>> config.setInitializeLogging(true);
<br>> config.setInitializeCache(true);
<br>> config.setInitializeReplication(true);
<br>> config.setErrorStream(System.err);
<br>> config.setMessageStream(System.out);
<br>> config.setReplicationTransport(id.intValue(), this);
<br>> dbEnv = new Environment(home, config);
<br>> dbEnv.checkpoint(null);
<br>> dbEnv.removeOldLogFiles();
<br>>
<br>> XmlManagerConfig xmlConfig = new XmlManagerConfig();
<br>> xmlConfig.setAdoptEnvironment(true);
<br>> xmlConfig.setAllowAutoOpen(true);
<br>> xmlConfig.setAllowExternalAccess(true);
<br>>
<br>> // Boolean used to know whether to close the environment
<br>> // when the cleanup() method is called.
<br>> this.dbEnvIsOpen = true;
<br>> this.mgr = new XmlManager(dbEnv, xmlConfig);
<br>> }
<br>> catch(Exception e)
<br>> {
<br>> e.printStackTrace(System.err);
<br>> throw new ExceptionInInitializerError(e.getMessage());
<br>> }
<br>> }
<br>>
<br>> // Returns the path to the database environment
<br>> public File getDbEnvPath()
<br>> {
<br>> return home;
<br>> }
<br>>
<br>> // Returns the database environment encapsulated by this class.
<br>> public Environment getEnvironment()
<br>> {
<br>> return dbEnv;
<br>> }
<br>>
<br>> // Returns the XmlManager encapsulated by this class.
<br>> public XmlManager getManager()
<br>> {
<br>> return mgr;
<br>> }
<br>>
<br>> // Used to close the environment
<br>> public void cleanup() throws DatabaseException
<br>> {
<br>> if (dbEnvIsOpen)
<br>> {
<br>> dbEnv.close();
<br>> dbEnvIsOpen = false;
<br>> }
<br>> }
<br>>
<br>> public int send(Environment dbenv, DatabaseEntry control,
<br>> DatabaseEntry rec, LogSequenceNumber lsn, int envid, boolean
<br>> noBuffer, boolean permanent) throws DatabaseException
<br>> {
<br>> int result = 0;
<br>> try
<br>> {
<br>> MyMsg mymsg = new MyMsg(control, rec, envid);
<br>> Message msg = new Message(null, null,
<br>> Util.objectToByteBuffer(mymsg));
<br>> channel.send(msg);
<br>> }
<br>> catch (Exception e)
<br>> {
<br>> result = -1;
<br>> }
<br>> return result;
<br>> }
<br>>
<br>> public void receive(Message message)
<br>> {
<br>> try
<br>> {
<br>> Object o= Util.objectToByteBuffer(message.getObject());
<br>> if(o instanceof MyMsg)
<br>> {
<br>> MyMsg mymsg = (MyMsg) o;
<br>> dbEnv.processReplicationMessage(mymsg.control,
<br>> mymsg.rec, mymsg.envid);
<br>> }
<br>> System.out.println(o + " [" + message.getSrc() + "]\n");
<br>> history.add(o);
<br>> }
<br>> catch(Exception e)
<br>> {
<br>> System.err.println("Chat.receive(): " + e);
<br>> }
<br>> return;
<br>> }
<br>>
<br>> public byte[] getState()
<br>> {
<br>> try
<br>> {
<br>> return Util.objectToByteBuffer(history);
<br>> }
<br>> catch(Exception e)
<br>> {
<br>> return null;
<br>> }
<br>> }
<br>>
<br>> public void setState(byte[] state)
<br>> {
<br>> try
<br>> {
<br>> history=(LinkedList)Util.objectFromByteBuffer(state);
<br>> }
<br>> catch(Exception e)
<br>> {
<br>> e.printStackTrace();
<br>> }
<br>> return;
<br>> }
<br>>
<br>> public void viewAccepted(View new_view)
<br>> {
<br>> System.out.println("Received view " + new_view + '\n');
<br>> int members = channel.getView().getMembers().size();
<br>>
<br>> if(members > 1)
<br>> {
<br>> System.out.println("We're now a group of " + members +
<br>> " members!");
<br>> try
<br>> {
<br>> Thread t = new Thread()
<br>> {
<br>> public void run()
<br>> {
<br>> try
<br>> {
<br>> int i = dbEnv.electReplicationMaster
<br>> (channel.getView().getMembers().size(), 2, 1, 1000);
<br>> masterId = new Integer(i);
<br>> }
<br>> catch (Exception e)
<br>> {
<br>> e.printStackTrace();
<br>> }
<br>> }
<br>> };
<br>>
<br>> t.start();
<br>> t.join();
<br>>
<br>> if(id.equals(masterId))
<br>> dbEnv.startReplication(new DatabaseEntry(), true);
<br>> else if(masterId.intValue() >= 0)
<br>> dbEnv.startReplication(new DatabaseEntry(),
<br>> false);
<br>> }
<br>> catch (Exception e)
<br>> {
<br>> e.printStackTrace();
<br>> }
<br>> }
<br>> return;
<br>> }
<br>>
<br>>
<br>> public void suspect(Address suspected_mbr)
<br>> {
<br>> }
<br>>
<br>>
<br>> public void block()
<br>> {
<br>> }
<br>>
<br>> class MyMsg implements Serializable
<br>> {
<br>>
<br>> public MyMsg(DatabaseEntry control, DatabaseEntry rec, int
<br>> envid)
<br>> {
<br>> this.control = control;
<br>> this.rec = rec;
<br>> this.envid = envid;
<br>> }
<br>> int envid;
<br>> DatabaseEntry control;
<br>> DatabaseEntry rec;
<br>> }
<br>>
<br>> class MyElectionThread implements Runnable
<br>> {
<br>> public void run()
<br>> {
<br>> try
<br>> {
<br>> }
<br>> catch (Exception e)
<br>> {
<br>> throw new RuntimeException(e.getMessage());
<br>> }
<br>> }
<br>> }
<br>> }
</div><br><br><br>------------------------------------------
<br>To remove yourself from this list, send an
<br>email to <a href="http://old.nabble.com/user/SendEmail.jtp?type=post&post=5594270&i=1" target="_top" rel="nofollow">xml-unsubscribe@...</a>
<br><br><p>From forum: <a href="http://old.nabble.com/Berkeley-DB-Xml---General-f732.html" embed="fixTarget[732]" target="_top" >Berkeley DB Xml - General</a></p>