|
View:
New views
20 Messages
—
Rating Filter:
Alert me
|
| < Prev | 1 - 2 | Next > |
|
|
CtrlClnt local/remote diff in return value on w2kHello Davide
Context : Server : W2K server SP4, xmail 1.26 WorkStation : W2K workstation SP4, ctrlclnt.exe copied from xmail 1.26 server (so exactly same binary) When running CtrlClnt from the workstation all ok, output result is ok, ctrlclnt exit immediatly and exit code (errorlevel) is 0 Same exactly command localy on the server, I get the same good output result , then ctrlclnt seems blocked for some seconds then finaly end but with exit code 1073807364 (decimal) It is not a urgent problem as remote is ok, and I manage 99%xmail remotely but wanted to tell you about this problem The only difference in result is the return code, making ctrlclnt usage impossible in batchs at server level (like some dump process, ...) :( Didn't try with a w2003 nor w2008 at server level (perhabs some having xmail on w2003 or w2008 could test and post results ?) Francis _______________________________________________ xmail mailing list xmail@... http://xmailserver.org/mailman/listinfo/xmail |
|
|
Re: CtrlClnt local/remote diff in return value on w2kAt 10.34 25/09/09, CLEMENT Francis wrote:
>Server : W2K server SP4, xmail 1.26 >WorkStation : W2K workstation SP4, ctrlclnt.exe copied from xmail 1.26 >server (so exactly same binary) > >When running CtrlClnt from the workstation all ok, output result is ok, >ctrlclnt exit immediatly and exit code (errorlevel) is 0 >Same exactly command localy on the server, I get the same good output result >, then ctrlclnt seems blocked for some seconds then finaly end but with exit >code 1073807364 (decimal) I run CtrlClnt locally on a XP Pro machine for changing passwords from a Perl script and it returns 0 on success. Connects to the server's public IP, though, not localhost. Things to consider: - different behaviour between WS and Server Windows versions (or XP vs 2K); - different behaviour when connecting to the public or the loopback IP I'm using 1.26, but I think the CtrlClnt.exe binary hasn't changed since 1.24. Ciao, Francesco _______________________________________________ xmail mailing list xmail@... http://xmailserver.org/mailman/listinfo/xmail |
|
|
|
|
|
|
|
|
|
|
|
Re: CtrlClnt local/remote diff in return value on w2kOn Fri, 25 Sep 2009, CLEMENT Francis wrote:
> Hello Davide > > Context : > Server : W2K server SP4, xmail 1.26 > WorkStation : W2K workstation SP4, ctrlclnt.exe copied from xmail 1.26 > server (so exactly same binary) > > When running CtrlClnt from the workstation all ok, output result is ok, > ctrlclnt exit immediatly and exit code (errorlevel) is 0 > Same exactly command localy on the server, I get the same good output result > , then ctrlclnt seems blocked for some seconds then finaly end but with exit > code 1073807364 (decimal) > > It is not a urgent problem as remote is ok, and I manage 99%xmail remotely > but wanted to tell you about this problem > The only difference in result is the return code, making ctrlclnt usage > impossible in batchs at server level (like some dump process, ...) :( > > Didn't try with a w2003 nor w2008 at server level (perhabs some having xmail > on w2003 or w2008 could test and post results ?) Using SSL or clear? - Davide _______________________________________________ xmail mailing list xmail@... http://xmailserver.org/mailman/listinfo/xmail |
|
|
Re: CtrlClnt local/remote diff in return value on w2kOn Fri, 25 Sep 2009, CLEMENT Francis wrote:
> >-----Message d'origine----- > >De : xmail-bounces@... > >[mailto:xmail-bounces@...]De la part de CLEMENT Francis > >Envoyé : vendredi 25 septembre 2009 14:14 > >À : 'XMail Users Mailing List' > >Objet : Re: [xmail] CtrlClnt local/remote diff in return value on w2k > > > > > >>-----Message d'origine----- > >>De : xmail-bounces@... > >>[mailto:xmail-bounces@...]De la part de Francesco Vertova > >>Envoyé : vendredi 25 septembre 2009 10:58 > >>À : XMail Users Mailing List > >>Objet : Re: [xmail] CtrlClnt local/remote diff in return value on w2k > >> > >> > >.... > > > >>I'm using 1.26, but I think the CtrlClnt.exe binary hasn't > >>changed since 1.24. > >> > >>Ciao, Francesco > >> > > > >I know that the problem is started with a 1.2x version but > >have to test to > >confirm (I keep olds in case :) ) > >I will test with previous 1.2x version of ctrlclnt.exe and report :) > > > >Francis > > > > Test with ctrlclnt version 1.26, 1.25 and 1.24 localy on server running > xmail server 1.26 (same ctrlclnt cmd line, same ctrl ip/port and action) > > 1.26 : all ok except delay 30s return code = 1073807364 > 1.25 : all ok except delay 30s return code = 1073807364 > 1.24 : all ok except delay 30s return code = 1073807364 > > Same versions on remote station : all ok, no delay, returns 0 - Davide _______________________________________________ xmail mailing list xmail@... http://xmailserver.org/mailman/listinfo/xmail |
|
|
|
|
|
|
|
|
Re: CtrlClnt local/remote diff in return value on w2kOn Mon, 28 Sep 2009, CLEMENT Francis wrote:
> >> Test with ctrlclnt version 1.26, 1.25 and 1.24 localy on > >server running > >> xmail server 1.26 (same ctrlclnt cmd line, same ctrl ip/port > >and action) > >> > >> 1.26 : all ok except delay 30s return code = 1073807364 > >> 1.25 : all ok except delay 30s return code = 1073807364 > >> 1.24 : all ok except delay 30s return code = 1073807364 > >> > >> Same versions on remote station : all ok, no delay, returns 0 > > > >What do you mean for "return code"? The CtrlClnt process exit code? > > Yep, ctrlclnt process exit code that you can test in batchs with errorlevel > variable on win systems. Hmmm ... CtrlClnt never returns such high error codes. Does it crash? Looks like some sort of error code returned by the system under certain abnormal conditions ... - Davide _______________________________________________ xmail mailing list xmail@... http://xmailserver.org/mailman/listinfo/xmail |
|
|
|
|
|
Re: CtrlClnt local/remote diff in return value on w2kOn Mon, 28 Sep 2009, CLEMENT Francis wrote:
> As said, ctrlclnt don't crash at all, at end of job (action correctly > completed at xmail server side, and ctrlclnt output correct and complete, > nothing missing), it seems to 'freeze' for about 30 seconds (waiting for a > timeout ?) before ending then returns normally (no crash dump, no drwatson, > no system event in logs, ...) but returns this 'strange' exit code. > > Same command from a workstation, do the same job, don't freeze at all after > final output result, and return 0 > > I need to test older ctrlclnt versions to find last that work 'localy' with > exit code 0 Looks like something triggered the debugger: http://www.ureader.com/message/1261932.aspx The error is DBG_TERMINATE_PROCESS. There are quite a few reports, if you google it. - Davide _______________________________________________ xmail mailing list xmail@... http://xmailserver.org/mailman/listinfo/xmail |
|
|
Re: CtrlClnt local/remote diff in return value on w2kOn Mon, 28 Sep 2009, Davide Libenzi wrote:
> On Mon, 28 Sep 2009, CLEMENT Francis wrote: > > > As said, ctrlclnt don't crash at all, at end of job (action correctly > > completed at xmail server side, and ctrlclnt output correct and complete, > > nothing missing), it seems to 'freeze' for about 30 seconds (waiting for a > > timeout ?) before ending then returns normally (no crash dump, no drwatson, > > no system event in logs, ...) but returns this 'strange' exit code. > > > > Same command from a workstation, do the same job, don't freeze at all after > > final output result, and return 0 > > > > I need to test older ctrlclnt versions to find last that work 'localy' with > > exit code 0 > > Looks like something triggered the debugger: > > http://www.ureader.com/message/1261932.aspx > > The error is DBG_TERMINATE_PROCESS. > There are quite a few reports, if you google it. If you install WinDBG and you enable just in time debugging, it should bring you right where the violation is. Because, this is very likely an error triggered by CtrlClnt. - Davide _______________________________________________ xmail mailing list xmail@... http://xmailserver.org/mailman/listinfo/xmail |
|
|
|
|
|
|
|
|
Re: CtrlClnt local/remote diff in return value on w2kOn Tue, 29 Sep 2009, CLEMENT Francis wrote:
> Here is WinDBG result (some change in cmd line to mask sensitive data) : > > CommandLine: C:\XMail\Bin\CtrlClnt.exe -s www.xxx.yyy.zzz -n 6017 -u > myusername -p mypassword userlist somehandleddomainname > Starting directory: C:\XMail\Bin > Symbol search path is: > SRV*c:\websymbols*http://msdl.microsoft.com/download/symbols > Executable search path is: > ModLoad: 00400000 0047e000 CtrlClnt.exe > ModLoad: 78460000 784e2000 ntdll.dll > ModLoad: 10000000 100fd000 C:\XMail\Bin\LIBEAY32.dll > ModLoad: 74fd0000 74fd9000 C:\WINNT\system32\WSOCK32.dll > ModLoad: 77e70000 77f2f000 C:\WINNT\system32\KERNEL32.DLL > ModLoad: 74fb0000 74fc4000 C:\WINNT\system32\WS2_32.DLL > ModLoad: 78000000 78045000 C:\WINNT\system32\MSVCRT.DLL > ModLoad: 78ed0000 78f35000 C:\WINNT\system32\ADVAPI32.DLL > ModLoad: 770c0000 7712f000 C:\WINNT\system32\RPCRT4.dll > ModLoad: 74fa0000 74fa8000 C:\WINNT\system32\WS2HELP.DLL > ModLoad: 77f40000 77f7d000 C:\WINNT\system32\GDI32.dll > ModLoad: 77e00000 77e5f000 C:\WINNT\system32\USER32.dll > ModLoad: 78130000 781cb000 C:\WINNT\system32\MSVCR80.dll > ModLoad: 00230000 00261000 C:\XMail\Bin\SSLEAY32.dll > (ad8.aac): Break instruction exception - code 80000003 (first chance) > eax=00000000 ebx=00131f04 ecx=00000000 edx=00132c48 esi=7ffdf000 > edi=00131f78 > eip=7847193c esp=0012f984 ebp=0012fc98 iopl=0 nv up ei pl nz na po > nc > cs=001b ss=0023 ds=0023 es=0023 fs=003b gs=0000 > efl=00000202 > ntdll!DbgBreakPoint: > 7847193c cc int 3 > 0:000> g > ModLoad: 74f50000 74f6d000 C:\WINNT\system32\msafd.dll > ModLoad: 74f90000 74f97000 C:\WINNT\System32\wshtcpip.dll > (ad8.888): Access violation - code c0000005 (first chance) > First chance exceptions are reported before any exception handling. > This exception may be expected and handled. > eax=00000000 ebx=00000000 ecx=0047a410 edx=00000ae4 esi=0047a400 > edi=00000000 > eip=7846f281 esp=00b2fba0 ebp=00b2fc00 iopl=0 nv up ei pl zr na pe > nc > cs=001b ss=0023 ds=0023 es=0023 fs=0038 gs=0000 > efl=00010246 > ntdll!RtlpWaitForCriticalSection+0x8a: > 7846f281 ff4010 inc dword ptr [eax+10h] > ds:0023:00000010=???????? > > I'm not a windbg specialist ;) What do you mean with "enable just in time > debugging" ? > Should I do somethink more with Windbg at this point ? Thank you Francis. Do you have a debug version of CtrlClnt to run (like the ones included in the recent -pre releases)? If you run the CtrlClnt and you have the corresponding PDB file in the same directory, WinDBG should resolve the symbols. Also, can you show a Call Stack (in the View menu) from the point of fault? - Davide _______________________________________________ xmail mailing list xmail@... http://xmailserver.org/mailman/listinfo/xmail |
|
|
|
|
|
SBS 2008 pop3connectorIs someone else having problem with this odd thing called SBS 2008
pop3connector? Many clients that use this connector are telling us that, for some reason, it is considering XMail´s headers as invalid. Some information from Microsoft´s site: Email Rejected Due to Protocol Errors (Invalid Headers) Note: A long-term solution to this issue is being investigated [New Content courtesy of Dan Thompson] Exchange will close an SMTP connection after a certain number of protocol errors (5 by default). (see the MaxProtocolErrors property of the ReceiveConnector object at: http://technet.microsoft.com/en-us/library/aa998618.aspx) When the SBS 2008 pop3connector downloads a message from a POP3 mailbox, it needs to figure out what the "return path" for the mail should be, which it does by reading the email's headers. The pop3connector does not do validation of the header value--it lets Exchange take care of that. If the header value that the pop3connector chooses is malformed, when it is sent to the Exchange server (as part of the "MAIL FROM" command), Exchange will reject it with a 501 error. That counts as a "protocol error", and therefore is counted against the MaxProtocolErrors limit. Since the pop3connector was not able to deliver the mail, and does not know if the mail is safe to delete, it leaves the mail on the POP3 server. If there are 5 of these messages in your POP3 mailbox, then there will be 5 "protocol errors" in the pop3connector's SMTP session, which hits the limit, and Exchange will end the session with a transient error (4xx). When this happens, the pop3connector recognizes that the error is transient, and will retry again at the next scheduled download period. But since those 5 malformed messages are still in the POP3 mailbox, the same thing will continue to happen, with no "forward progress" being made. The workaround is to increase the "MaxProtocolErrors" property of the Windows SBS Fax Sharepoint Receive connector, and then restart the Exchange Transport service for the change to take effect (and you'll have to restart the pop3connector service, too, since it depends on the Exchange Transport service). Unfortunately, you can't set that property from the Exchange management GUI, so you have to do it from an (elevated) Exchange Powershell prompt. Here are the instructions: >From an elevated Exchange Management Shell (Exchange Powershell window) (right click on "Start-->Microsoft Exchange Server 2007-->Exchange Management Shell" and then choose "Run as administrator") run the following Powershell commands: Set-ReceiveConnector -Identity ($Env:computername + "\Windows SBS Fax Sharepoint Receive " + $Env:computername) -MaxProtocolErrors 500 Stop-Service pop3connector Restart-Service -force MSExchangeTransportStart-Service pop3connector That will increase the MaxProtocol errors (of the internal receive connector only) to match the pop3connector's max emails downloaded per session. Once you get 500 messages with malformed headers stacked up in the POP3 mailbox, though, you'll still have to delete them manually. Regards Edinilson ------------------------------------------------:--------- ATINET-Professional Web Hosting Tel Voz: (0xx11) 4412-0876 http://www.atinet.com.br _______________________________________________ xmail mailing list xmail@... http://xmailserver.org/mailman/listinfo/xmail |
|
|
Re: SBS 2008 pop3connectorOn Wed, 30 Sep 2009, Edinilson - ATINET wrote:
> Is someone else having problem with this odd thing called SBS 2008 > pop3connector? > Many clients that use this connector are telling us that, for some reason, > it is considering XMail´s headers as invalid. No idea. It'd be great to know what are those errors (or better, that MS considers so). - Davide _______________________________________________ xmail mailing list xmail@... http://xmailserver.org/mailman/listinfo/xmail |
|
|
Re: CtrlClnt local/remote diff in return value on w2kOn Wed, 30 Sep 2009, CLEMENT Francis wrote:
> Here is it : > Debug with ctrlclnt v 1.26pre05, Call Stack (with source and source args > options activated) : > > ntdll!RtlpWaitForCriticalSection+0x8a > ntdll!RtlEnterCriticalSection+0x46 > CtrlClnt!SysCreateTlsKey(int * TlsKey = 0x0047a3cc, <function> * pFreeProc = > 0x00404c60)+0xf [c:\temp\xmail-1.26-pre05\sysdepwin.cpp @ 1216] > CtrlClnt!ErrOnceSetup(void)+0x15 [c:\temp\xmail-1.26-pre05\errors.cpp @ 333] > CtrlClnt!SysThreadOnce(struct SYS_THREAD_ONCE * pThrOnce = 0x0047a3d0, > <function> * pOnceProc = 0x00404bf0)+0x16 > [c:\temp\xmail-1.26-pre05\sysdepwin.cpp @ 1287] > CtrlClnt!ErrSetupEnv(void)+0x15 [c:\temp\xmail-1.26-pre05\errors.cpp @ 343] > CtrlClnt!ErrSetErrorCode(int iError = -5, char * pszInfo = 0x00000000 "", > int iSize = -1)+0x9 [c:\temp\xmail-1.26-pre05\errors.cpp @ 380] > CtrlClnt!SysLockMutex(void * hMutex = 0xfeeefeee, int iTimeout = > 4194304)+0x26 [c:\temp\xmail-1.26-pre05\sysdepwin.cpp @ 937] > CtrlClnt!BSslLockingCB(int iMode = 9, int iType = 1, char * pszFile = > 0x100b2a44 ".\crypto\err\err.c", int iLine = 419)+0x22 > [c:\temp\xmail-1.26-pre05\sslbind.cpp @ 82] > WARNING: Stack unwind information not available. Following frames may be > wrong. > LIBEAY32!CRYPTO_lock+0x81 > LIBEAY32!ENGINE_load_ubsec+0x7bb > ntdll!LdrShutdownThread+0xa3 > KERNEL32!ExitThread+0x53 > KERNEL32!FreeLibraryAndExitThread+0x24 > msafd!SockAsyncThread+0xe7 > KERNEL32!BaseThreadStart+0x52 Great trace. Need to investigate ... - Davide _______________________________________________ xmail mailing list xmail@... http://xmailserver.org/mailman/listinfo/xmail |
| < Prev | 1 - 2 | Next > |
| Free embeddable forum powered by Nabble | Forum Help |