Re: [PATCHES] PostgreSQL 8.0.3 and Ipv6

[Andrew Dunstan]

[moving lists]

First, this question should not be on the -patches list. Please read the
mailing list descriptions carefully before posting and use the right one
- in this case probably the -hackers-win32 list.

IPv6 is a build time setting, and it appears that the current build
environment used for Windows (or at least mine) does not support IPv6.
It looks for struct sockaddr_in6 and fails to find it, apparently.

In fact, I am surprised that your postmaster doesn't complain if you put
an IPv6 address in pg_hba.conf.

cheers

andrew



Solange wrote:

> Hi, I am with problems to function postgreSQL 8.0.3 in
> windows XP with the Ipv6 protocol. I try to have access the
> bank in the following way:
>
>  psql -h ::1 -d db_teste -u postgres
>
>  but it gives an error: could not translate ::1: Unknown host.
>  already I configured the archive pg_hba.conf so that it has
> accepted address::1 of ipv6. My machine is effecting ping6 in
> ::1 and my apache server also is functioning in ipv6. I made
> everything what it said in the manual.
>  
> My question is the following one: PostgreSQL does not function in
> windows XP
> with the Ipv6 protocol, or has that to configure plus some
> thing? Or It only functions in linux?
> Necessary of some patch so that it functions in windows Xp?

---------------------------(end of broadcast)---------------------------
TIP 6: explain analyze is your friend

[Andrew Dunstan]

Researching this some more, I see:

/mingw/include/ws2tcpip.h:struct sockaddr_in6 {

do we need to include this somewhere?

cheers

andrew

Andrew Dunstan said:

>
> [moving lists]
>
> First, this question should not be on the -patches list. Please read
> the  mailing list descriptions carefully before posting and use the
> right one  - in this case probably the -hackers-win32 list.
>
> IPv6 is a build time setting, and it appears that the current build
> environment used for Windows (or at least mine) does not support IPv6.
> It looks for struct sockaddr_in6 and fails to find it, apparently.
>
> In fact, I am surprised that your postmaster doesn't complain if you
> put  an IPv6 address in pg_hba.conf.
>
> cheers
>
> andrew
>
>
>
> Solange wrote:
>
>> Hi, I am with problems to function postgreSQL 8.0.3 in
>> windows XP with the Ipv6 protocol. I try to have access the
>> bank in the following way:
>>
>>  psql -h ::1 -d db_teste -u postgres
>>
>>  but it gives an error: could not translate ::1: Unknown host.
>>  already I configured the archive pg_hba.conf so that it has
>> accepted address::1 of ipv6. My machine is effecting ping6 in
>> ::1 and my apache server also is functioning in ipv6. I made
>> everything what it said in the manual.
>>
>> My question is the following one: PostgreSQL does not function in
>> windows XP
>> with the Ipv6 protocol, or has that to configure plus some
>> thing? Or It only functions in linux?
>> Necessary of some patch so that it functions in windows Xp?
>
>
> ---------------------------(end of
> broadcast)--------------------------- TIP 6: explain analyze is your
> friend

---------------------------(end of broadcast)---------------------------
TIP 6: explain analyze is your friend

[Magnus Hagander]

> Researching this some more, I see:
>
> /mingw/include/ws2tcpip.h:struct sockaddr_in6 {
>
> do we need to include this somewhere?

Probably.
Note that most of what's in that file is only supported on win2000 and
newer, IIRC. So I'm unsure wether we can include it unconditionally - it
has to be tested.

//Magnus

---------------------------(end of broadcast)---------------------------
TIP 9: In versions below 8.0, the planner will ignore your desire to
       choose an index scan if your joining column's datatypes do not
       match

[Andrew Dunstan]

Magnus Hagander wrote:

>>Researching this some more, I see:
>>
>>/mingw/include/ws2tcpip.h:struct sockaddr_in6 {
>>
>>do we need to include this somewhere?
>>    
>>
>
>Probably.
>Note that most of what's in that file is only supported on win2000 and
>newer, IIRC. So I'm unsure wether we can include it unconditionally - it
>has to be tested.
>
>
>  
>

Ouch! Remind me again why we decided to support NT4?

Does anyone have an NT4 build enviroment that they can test patches on
if I come up with a fix? I sure don't.

cheers

andrew

---------------------------(end of broadcast)---------------------------
TIP 3: Have you checked our extensive FAQ?

               http://www.postgresql.org/docs/faq

[Petr Jelinek-2]

Andrew Dunstan wrote:
>
> Ouch! Remind me again why we decided to support NT4?
>
> Does anyone have an NT4 build enviroment that they can test patches on
> if I come up with a fix? I sure don't.

The truth is that IPv6 isn't even in Windows 2000.
There is some "technology preview" ipv6 implementation for Windows 2000
which could be downloaded from MS website but it should not be used in
production enviroments.

I can test patches on my W2k if needed (I mean I can test if Postgres
still works without IPv6 support after patching because I don't have IPv6).

--
Regards
Petr Jelinek (PJMODOS)

---------------------------(end of broadcast)---------------------------
TIP 6: explain analyze is your friend

[Magnus Hagander]

> >>Researching this some more, I see:
> >>
> >>/mingw/include/ws2tcpip.h:struct sockaddr_in6 {
> >>
> >>do we need to include this somewhere?
> >>    
> >>
> >
> >Probably.
> >Note that most of what's in that file is only supported on
> win2000 and
> >newer, IIRC. So I'm unsure wether we can include it
> unconditionally -
> >it has to be tested.
> >
> >
> >  
> >
>
> Ouch! Remind me again why we decided to support NT4?

Because people use it?
We don's supoprt it in the installer, but IIRC we've had a couple of
requests for it. So *somebody* uses it.

I haven't looked at things, but I assume client code is also affected in
some ways? If so, we also support 9x.


> Does anyone have an NT4 build enviroment that they can test
> patches on if I come up with a fix? I sure don't.

Build, no. Runtime, yes. I keep a couple of vmwares with NT4 around to
do this kind of testing on.

I'd be fine with requiring 2000+ to build as long as it can be *used* on
NT4...

//Magnus

---------------------------(end of broadcast)---------------------------
TIP 5: don't forget to increase your free space map settings

[Andrew Dunstan]

Magnus Hagander wrote:

>>Does anyone have an NT4 build enviroment that they can test
>>patches on if I come up with a fix? I sure don't.
>>    
>>
>
>Build, no. Runtime, yes. I keep a couple of vmwares with NT4 around to
>do this kind of testing on.
>
>I'd be fine with requiring 2000+ to build as long as it can be *used* on
>NT4...
>
>
>  
>

If I manage to get anything built I will get you the binaries to test with.

Right now, having faked out the configure test for getaddrinfo, I am
snookered by a mingw bug that has gai_strerror defined to a non-existent
object (In the MS headers it is defined inline in ws2tcpip.h). See
https://sourceforge.net/tracker/?func=detail&atid=102435&aid=1158095&group_id=2435

This has linking for postmaster failing with unresolved symbol errors.

What a mess. ;-(

cheers

andrew



---------------------------(end of broadcast)---------------------------
TIP 5: don't forget to increase your free space map settings

[Joshua D. Drake]

>Build, no. Runtime, yes. I keep a couple of vmwares with NT4 around to
>do this kind of testing on.
>
>I'd be fine with requiring 2000+ to build as long as it can be *used* on
>NT4...
>  
>
NT4 is dead. Let it die. It is very, very old and no longer supported
by Microsoft.

2000+ should be the minimum.

Just curious but would PostgreSQL compile on RedHat 6.2 and run?

Sincerely,

Joshua D. Drake



>//Magnus
>
>---------------------------(end of broadcast)---------------------------
>TIP 5: don't forget to increase your free space map settings
>  
>


---------------------------(end of broadcast)---------------------------
TIP 1: if posting/reading through Usenet, please send an appropriate
       subscribe-nomail command to majordomo@... so that your
       message can get through to the mailing list cleanly

[Merlin Moncure]

Joshua wrote:
> NT4 is dead. Let it die. It is very, very old and no longer supported
> by Microsoft.
>
> 2000+ should be the minimum.
>
> Just curious but would PostgreSQL compile on RedHat 6.2 and run?
>
Windows NT 4.0 was released in Jul of 1996.  The official release of red
had linux at that time was 3.04 with the upcoming 4.0 in beta.

Just to put things into perspective :).

(I agree with you, let it die!)

Merlin

---------------------------(end of broadcast)---------------------------
TIP 2: Don't 'kill -9' the postmaster

[Andrew Dunstan]

Merlin Moncure wrote:

>Joshua wrote:
>  
>
>>NT4 is dead. Let it die. It is very, very old and no longer supported
>>by Microsoft.
>>
>>2000+ should be the minimum.
>>
>>Just curious but would PostgreSQL compile on RedHat 6.2 and run?
>>
>>    
>>
>Windows NT 4.0 was released in Jul of 1996.  The official release of red
>had linux at that time was 3.04 with the upcoming 4.0 in beta.
>
>Just to put things into perspective :).
>
>
>  
>

All this is irrelevant to the present discussion, as W2K does not
support what I tried to do.

In any case, EOL date matters far more than release date, and how many
users there still are matters even more than either date. RH7.3 has been
EOLed for some time, but I see huge amounts of it still in data centers.


cheers

andrew

---------------------------(end of broadcast)---------------------------
TIP 4: Have you searched our list archives?

               http://archives.postgresql.org

[Dave Page]

 

> -----Original Message-----
> From: pgsql-hackers-win32-owner@...
> [mailto:pgsql-hackers-win32-owner@...] On Behalf
> Of Merlin Moncure
> Sent: 19 August 2005 19:03
> To: Joshua D. Drake
> Cc: pgsql-hackers-win32@...
> Subject: Re: [pgsql-hackers-win32] [PATCHES] PostgreSQL 8.0.3 and Ipv6
>
> Joshua wrote:
> > NT4 is dead. Let it die. It is very, very old and no longer
> supported
> > by Microsoft.
> >
> > 2000+ should be the minimum.
> >
> > Just curious but would PostgreSQL compile on RedHat 6.2 and run?
> >
> Windows NT 4.0 was released in Jul of 1996.  The official
> release of red
> had linux at that time was 3.04 with the upcoming 4.0 in beta.
>
> Just to put things into perspective :).
>
> (I agree with you, let it die!)

+1

FWIW, pgInstaller won't run fully on NT because we can't initdb as a
different user (the APIs simply aren't there in <2K).

/D

---------------------------(end of broadcast)---------------------------
TIP 1: if posting/reading through Usenet, please send an appropriate
       subscribe-nomail command to majordomo@... so that your
       message can get through to the mailing list cleanly