Interesting streaming replication issue

(sorry for the top post, bitchy K9 Mail)

James,

are you sure you're scp'ing from the archive, not from pg_xlog?

Regards,

Gunnar "Nick" Bluth

Am 27. Juli 2017 05:00:17 MESZ schrieb James Sewell <james.sewell@jirotech.com>:

Hi all,

I've got two servers (A,B) which are part of a streaming replication
pair.
A is the master, B is a hot standby. I'm sending archived WAL to a
directory on A, B is reading it via SCP.

This all works fine normally. I'm on Redhat 7.3, running EDB 9.6.2 (I'm
currently working to reproduce with standard 9.6)

We have recently seen a situation where B does not catch up when taken
offline for maintenance.

When B is started we see the following in the logs:

2017-07-27 11:55:57 AEST [21432]: [979-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000AA" from
archive
2017-07-27 11:55:58 AEST [21432]: [980-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000AB" from
archive
2017-07-27 11:55:58 AEST [21432]: [981-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000AC" from
archive
2017-07-27 11:55:59 AEST [21432]: [982-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000AD" from
archive
2017-07-27 11:55:59 AEST [21432]: [983-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000AE" from
archive
2017-07-27 11:56:00 AEST [21432]: [984-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000AF" from
archive
2017-07-27 11:56:00 AEST [21432]: [985-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000B0" from
archive
2017-07-27 11:56:01 AEST [21432]: [986-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000B1" from
archive
2017-07-27 11:56:01 AEST [21432]: [987-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000B2" from
archive
2017-07-27 11:56:02 AEST [21432]: [988-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000B3" from
archive
2017-07-27 11:56:02 AEST [21432]: [989-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000B4" from
archive
2017-07-27 11:56:03 AEST [21432]: [990-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000B5" from
archive
scp: /archive/xlog//0000000C0000005A000000B6: No such file or directory
2017-07-27 11:56:03 AEST [46191]: [1-1] user=,db=,client=
(0:00000)LOG: started streaming WAL from primary at 5A/B5000000 on
timeline 12
2017-07-27 11:56:03 AEST [46191]: [2-1] user=,db=,client=
(0:XX000)FATAL: could not receive data from WAL stream: ERROR:
requested WAL segment 0000000C0000005A000000B5 has already been
removed

scp: /archive/xlog//0000000D.history: No such file or directory
scp: /archive/xlog//0000000C0000005A000000B6: No such file or directory
2017-07-27 11:56:04 AEST [46203]: [1-1] user=,db=,client=
(0:00000)LOG: started streaming WAL from primary at 5A/B5000000 on
timeline 12
2017-07-27 11:56:04 AEST [46203]: [2-1] user=,db=,client=
(0:XX000)FATAL: could not receive data from WAL stream: ERROR:
requested WAL segment 0000000C0000005A000000B5 has already been
removed

This will loop indefinitely. At this stage the master reports no
connected
standbys in pg_stat_replication, and the standby has no running WAL
receiver process.

This can be 'fixed' by running pg_switch_xlog() on the master, at which
time a connection is seen from the standby and the logs show the
following:

scp: /archive/xlog//0000000D.history: No such file or directory
2017-07-27 12:03:19 AEST [21432]: [1029-1] user=,db=,client=
(0:00000)LOG:
restored log file "0000000C0000005A000000B5" from archive
scp: /archive/xlog//0000000C0000005A000000B6: No such file or directory
2017-07-27 12:03:19 AEST [63141]: [1-1] user=,db=,client=
(0:00000)LOG:
started streaming WAL from primary at 5A/B5000000 on timeline 12
2017-07-27 12:03:19 AEST [63141]: [2-1] user=,db=,client=
(0:XX000)FATAL:
could not receive data from WAL stream: ERROR: requested WAL segment
0000000C0000005A000000B5 has already been removed

scp: /archive/xlog//0000000D.history: No such file or directory
2017-07-27 12:03:24 AEST [21432]: [1030-1] user=,db=,client=
(0:00000)LOG:
restored log file "0000000C0000005A000000B5" from archive
2017-07-27 12:03:24 AEST [21432]: [1031-1] user=,db=,client=
(0:00000)LOG:
restored log file "0000000C0000005A000000B6" from archive
scp: /archive/xlog//0000000C0000005A000000B7: No such file or directory
2017-07-27 12:03:25 AEST [21432]: [1032-1] user=,db=,client=
(0:00000)LOG:
unexpected pageaddr 56/E7000000 in log segment
0000000C0000005A000000B7,
offset 0
2017-07-27 12:03:25 AEST [63196]: [1-1] user=,db=,client=
(0:00000)LOG:
started streaming WAL from primary

At this time replication is running as normal and all changes are
streamed.

Am I missing something here, this seems very odd. One thing I have
noticed
is it only seems to be caused after a lot of WAL is produced and the
pg_xlog
directory is sitting at max_wal_size

James Sewell,
PostgreSQL Team Lead / Solutions Architect

Suite 112, Jones Bay Wharf, 26-32 Pirrama Road, Pyrmont NSW 2009
*P *(+61) 2 8099 9000 <(+61)%202%208099%209000> *W* www.jirotech.com
*F *
(+61) 2 8099 9099 <(+61)%202%208099%209000>

--

------------------------------
The contents of this email are confidential and may be subject to legal
or
professional privilege and copyright. No representation is made that
this
email is free of viruses or other defects. If you have received this
communication in error, you may not copy or distribute any part of it
or
otherwise disclose its contents to anyone. Please advise the sender of
your
incorrect receipt of this correspondence.

--
Diese Nachricht wurde von meinem Android-Mobiltelefon mit K-9 Mail gesendet.

On Thu, Jul 27, 2017 at 4:41 PM, Gunnar "Nick" Bluth <
gunnar.bluth@pro-open.de> wrote:

are you sure you're scp'ing from the archive, not from pg_xlog?

Yes:

restore_command = 'scp -o StrictHostKeyChecking=no
10.154.19.30:/archive/xlog//%f
%p'

Although you are right - that would almost make sense if I had done that!

Cheers,
James

Regards,

Gunnar "Nick" Bluth

Am 27. Juli 2017 05:00:17 MESZ schrieb James Sewell <
james.sewell@jirotech.com>:

Hi all,

I've got two servers (A,B) which are part of a streaming replication
pair. A is the master, B is a hot standby. I'm sending archived WAL to a
directory on A, B is reading it via SCP.

This all works fine normally. I'm on Redhat 7.3, running EDB 9.6.2 (I'm
currently working to reproduce with standard 9.6)

We have recently seen a situation where B does not catch up when taken
offline for maintenance.

When B is started we see the following in the logs:

2017-07-27 11:55:57 AEST [21432]: [979-1] user=,db=,client= (0:00000)LOG: restored log file "0000000C0000005A000000AA" from archive
2017-07-27 11:55:58 AEST [21432]: [980-1] user=,db=,client= (0:00000)LOG: restored log file "0000000C0000005A000000AB" from archive
2017-07-27 11:55:58 AEST [21432]: [981-1] user=,db=,client= (0:00000)LOG: restored log file "0000000C0000005A000000AC" from archive
2017-07-27 11:55:59 AEST [21432]: [982-1] user=,db=,client= (0:00000)LOG: restored log file "0000000C0000005A000000AD" from archive
2017-07-27 11:55:59 AEST [21432]: [983-1] user=,db=,client= (0:00000)LOG: restored log file "0000000C0000005A000000AE" from archive
2017-07-27 11:56:00 AEST [21432]: [984-1] user=,db=,client= (0:00000)LOG: restored log file "0000000C0000005A000000AF" from archive
2017-07-27 11:56:00 AEST [21432]: [985-1] user=,db=,client= (0:00000)LOG: restored log file "0000000C0000005A000000B0" from archive
2017-07-27 11:56:01 AEST [21432]: [986-1] user=,db=,client= (0:00000)LOG: restored log file "0000000C0000005A000000B1" from archive
2017-07-27 11:56:01 AEST [21432]: [987-1] user=,db=,client= (0:00000)LOG: restored log file "0000000C0000005A000000B2" from archive
2017-07-27 11:56:02 AEST [21432]: [988-1] user=,db=,client= (0:00000)LOG: restored log file "0000000C0000005A000000B3" from archive
2017-07-27 11:56:02 AEST [21432]: [989-1] user=,db=,client= (0:00000)LOG: restored log file "0000000C0000005A000000B4" from archive
2017-07-27 11:56:03 AEST [21432]: [990-1] user=,db=,client= (0:00000)LOG: restored log file "0000000C0000005A000000B5" from archive
scp: /archive/xlog//0000000C0000005A000000B6: No such file or directory
2017-07-27 11:56:03 AEST [46191]: [1-1] user=,db=,client= (0:00000)LOG: started streaming WAL from primary at 5A/B5000000 on timeline 12
2017-07-27 11:56:03 AEST [46191]: [2-1] user=,db=,client= (0:XX000)FATAL: could not receive data from WAL stream: ERROR: requested WAL segment 0000000C0000005A000000B5 has already been removed

scp: /archive/xlog//0000000D.history: No such file or directory
scp: /archive/xlog//0000000C0000005A000000B6: No such file or directory
2017-07-27 11:56:04 AEST [46203]: [1-1] user=,db=,client= (0:00000)LOG: started streaming WAL from primary at 5A/B5000000 on timeline 12
2017-07-27 11:56:04 AEST [46203]: [2-1] user=,db=,client= (0:XX000)FATAL: could not receive data from WAL stream: ERROR: requested WAL segment 0000000C0000005A000000B5 has already been removed

This will loop indefinitely. At this stage the master reports no
connected standbys in pg_stat_replication, and the standby has no
running WAL receiver process.

This can be 'fixed' by running pg_switch_xlog() on the master, at which
time a connection is seen from the standby and the logs show the following:

scp: /archive/xlog//0000000D.history: No such file or directory
2017-07-27 12:03:19 AEST [21432]: [1029-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000B5" from archive
scp: /archive/xlog//0000000C0000005A000000B6: No such file or directory
2017-07-27 12:03:19 AEST [63141]: [1-1] user=,db=,client= (0:00000)LOG:
started streaming WAL from primary at 5A/B5000000 on timeline 12
2017-07-27 12:03:19 AEST [63141]: [2-1] user=,db=,client=
(0:XX000)FATAL: could not receive data from WAL stream: ERROR: requested
WAL segment 0000000C0000005A000000B5 has already been removed

scp: /archive/xlog//0000000D.history: No such file or directory
2017-07-27 12:03:24 AEST [21432]: [1030-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000B5" from archive
2017-07-27 12:03:24 AEST [21432]: [1031-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000B6" from archive
scp: /archive/xlog//0000000C0000005A000000B7: No such file or directory
2017-07-27 12:03:25 AEST [21432]: [1032-1] user=,db=,client=
(0:00000)LOG: unexpected pageaddr 56/E7000000 in log segment
0000000C0000005A000000B7, offset 0
2017-07-27 12:03:25 AEST [63196]: [1-1] user=,db=,client= (0:00000)LOG:
started streaming WAL from primary

At this time replication is running as normal and all changes are
streamed.

Am I missing something here, this seems very odd. One thing I have
noticed is it only seems to be caused after a lot of WAL is produced and
the pg_xlog directory is sitting at max_wal_size

James Sewell,
PostgreSQL Team Lead / Solutions Architect

Suite 112, Jones Bay Wharf, 26-32 Pirrama Road, Pyrmont NSW 2009
*P *(+61) 2 8099 9000 <(+61)%202%208099%209000> *W* www.jirotech.com
*F *(+61) 2 8099 9099 <(+61)%202%208099%209000>

------------------------------
The contents of this email are confidential and may be subject to legal
or professional privilege and copyright. No representation is made that
this email is free of viruses or other defects. If you have received this
communication in error, you may not copy or distribute any part of it or
otherwise disclose its contents to anyone. Please advise the sender of your
incorrect receipt of this correspondence.

--
Diese Nachricht wurde von meinem Android-Mobiltelefon mit K-9 Mail
gesendet.

--

------------------------------
The contents of this email are confidential and may be subject to legal or
professional privilege and copyright. No representation is made that this
email is free of viruses or other defects. If you have received this
communication in error, you may not copy or distribute any part of it or
otherwise disclose its contents to anyone. Please advise the sender of your
incorrect receipt of this correspondence.

On Wed, Jul 26, 2017 at 11:55 PM, James Sewell <james.sewell@jirotech.com>
wrote:

On Thu, Jul 27, 2017 at 4:41 PM, Gunnar "Nick" Bluth <
gunnar.bluth@pro-open.de> wrote:

are you sure you're scp'ing from the archive, not from pg_xlog?

Yes:

restore_command = 'scp -o StrictHostKeyChecking=no 10.154.19.30:/archive/xlog//%f
%p'

Although you are right - that would almost make sense if I had done that!

Sounds a lot like a cleanup process on your archive directory or something
getting in the way. Are the logs pg is asking for in that archive dir?

are you sure you're scp'ing from the archive, not from pg_xlog?

Yes:

restore_command = 'scp -o StrictHostKeyChecking=no 10.154.19.30:/archive/xlog//%f
%p'

Although you are right - that would almost make sense if I had done that!

Sounds a lot like a cleanup process on your archive directory or something
getting in the way. Are the logs pg is asking for in that archive dir?

That's the strange thing - if you look at the log not only are they there,
the standby has already retrieved them.

It's then asking for the log again via the stream.
--
James Sewell,
PostgreSQL Team Lead / Solutions Architect

Suite 112, Jones Bay Wharf, 26-32 Pirrama Road, Pyrmont NSW 2009
*P *(+61) 2 8099 9000 <javascript:void(0);> *W* www.jirotech.com *F *
(+61) 2 8099 9099 <javascript:void(0);>

--

------------------------------
The contents of this email are confidential and may be subject to legal or
professional privilege and copyright. No representation is made that this
email is free of viruses or other defects. If you have received this
communication in error, you may not copy or distribute any part of it or
otherwise disclose its contents to anyone. Please advise the sender of your
incorrect receipt of this correspondence.

OK this is reproducible now.

1. Stop a standby
2. Write some data to the master
3. Wait till the master has archived some WAL logs
4. Wait till the archived logs have been removed from pg_xlog
5. Start the standby.

The standby will recover all logs from the master log archive up to log X,
it will then try to get log X+1 and fail (doesn't exist).

It will then try to start streaming log X (not X+1) from the master and
fail (it's been archived). This will loop forever, example below.

scp: /archive/xlog//0000000D.history: No such file or directory
2017-08-03 10:26:41 AEST [578]: [1037-1] user=,db=,client= (0:00000)LOG:
restored log file "0000000C0000006 E000000AE" from archive
scp: /archive/xlog//0000000C0000006E000000AF: No such file or directory
2017-08-03 10:26:41 AEST [68161]: [1-1] user=,db=,client= (0:00000)LOG:
started streaming WAL from primary at 6E/AE000000 on timeline 12
2017-08-03 10:26:41 AEST [68161]: [2-1] user=,db=,client= (0:XX000)FATAL:
could not receive data from WAL s tream: ERROR: requested WAL segment
0000000C0000006E000000AE has already been removed

At this stage the standby has log X in pg_xlog, and this log has an
identical md5 checksum to the log in the master archive.

Performing a pg_switch_xlog on the master pushes log X+1 to the archive,
which is picked up by the standby allowing streaming replication to start.

The only interesting thing I can see in log X is that it's 99% made up
of FPI_FOR_HINT records.

Any ideas?

Cheers,
James

James Sewell,
PostgreSQL Team Lead / Solutions Architect

Suite 112, Jones Bay Wharf, 26-32 Pirrama Road, Pyrmont NSW 2009
*P *(+61) 2 8099 9000 <(+61)%202%208099%209000> *W* www.jirotech.com *F *
(+61) 2 8099 9099 <(+61)%202%208099%209000>

On Fri, Jul 28, 2017 at 6:28 AM, James Sewell <james.sewell@jirotech.com>
wrote:

are you sure you're scp'ing from the archive, not from pg_xlog?

Yes:

restore_command = 'scp -o StrictHostKeyChecking=no 10.154.19.30:/archive/xlog//%f
%p'

Although you are right - that would almost make sense if I had done that!

Sounds a lot like a cleanup process on your archive directory or
something getting in the way. Are the logs pg is asking for in that archive
dir?

That's the strange thing - if you look at the log not only are they there,
the standby has already retrieved them.

It's then asking for the log again via the stream.
--
James Sewell,
PostgreSQL Team Lead / Solutions Architect

Suite 112, Jones Bay Wharf, 26-32 Pirrama Road, Pyrmont NSW 2009
*P *(+61) 2 8099 9000 *W* www.jirotech.com *F *(+61) 2 8099 9099

--

------------------------------
The contents of this email are confidential and may be subject to legal or
professional privilege and copyright. No representation is made that this
email is free of viruses or other defects. If you have received this
communication in error, you may not copy or distribute any part of it or
otherwise disclose its contents to anyone. Please advise the sender of your
incorrect receipt of this correspondence.

Hi,

On 2017-07-27 13:00:17 +1000, James Sewell wrote:

Hi all,

I've got two servers (A,B) which are part of a streaming replication pair.
A is the master, B is a hot standby. I'm sending archived WAL to a
directory on A, B is reading it via SCP.

This all works fine normally. I'm on Redhat 7.3, running EDB 9.6.2 (I'm
currently working to reproduce with standard 9.6)

We have recently seen a situation where B does not catch up when taken
offline for maintenance.

When B is started we see the following in the logs:

2017-07-27 11:56:03 AEST [21432]: [990-1] user=,db=,client=
(0:00000)LOG: restored log file "0000000C0000005A000000B5" from
archive
scp: /archive/xlog//0000000C0000005A000000B6: No such file or directory
2017-07-27 11:56:03 AEST [46191]: [1-1] user=,db=,client=
(0:00000)LOG: started streaming WAL from primary at 5A/B5000000 on
timeline 12
2017-07-27 11:56:03 AEST [46191]: [2-1] user=,db=,client=
(0:XX000)FATAL: could not receive data from WAL stream: ERROR:
requested WAL segment 0000000C0000005A000000B5 has already been
removed

scp: /archive/xlog//0000000D.history: No such file or directory
scp: /archive/xlog//0000000C0000005A000000B6: No such file or directory
2017-07-27 11:56:04 AEST [46203]: [1-1] user=,db=,client=
(0:00000)LOG: started streaming WAL from primary at 5A/B5000000 on
timeline 12
2017-07-27 11:56:04 AEST [46203]: [2-1] user=,db=,client=
(0:XX000)FATAL: could not receive data from WAL stream: ERROR:
requested WAL segment 0000000C0000005A000000B5 has already been
removed

This will loop indefinitely. At this stage the master reports no connected
standbys in pg_stat_replication, and the standby has no running WAL
receiver process.

This can be 'fixed' by running pg_switch_xlog() on the master, at which
time a connection is seen from the standby and the logs show the following:

scp: /archive/xlog//0000000D.history: No such file or directory
2017-07-27 12:03:19 AEST [21432]: [1029-1] user=,db=,client= (0:00000)LOG:
restored log file "0000000C0000005A000000B5" from archive
scp: /archive/xlog//0000000C0000005A000000B6: No such file or directory
2017-07-27 12:03:19 AEST [63141]: [1-1] user=,db=,client= (0:00000)LOG:
started streaming WAL from primary at 5A/B5000000 on timeline 12
2017-07-27 12:03:19 AEST [63141]: [2-1] user=,db=,client= (0:XX000)FATAL:
could not receive data from WAL stream: ERROR: requested WAL segment
0000000C0000005A000000B5 has already been removed

scp: /archive/xlog//0000000D.history: No such file or directory
2017-07-27 12:03:24 AEST [21432]: [1030-1] user=,db=,client= (0:00000)LOG:
restored log file "0000000C0000005A000000B5" from archive
2017-07-27 12:03:24 AEST [21432]: [1031-1] user=,db=,client= (0:00000)LOG:
restored log file "0000000C0000005A000000B6" from archive

FWIW, I don't see a bug here. Archiving on its own doesn't guarantee
that replication progresses in increments smaller than 16MB, unless you
use archive_timeout (or as you do manually switch segments). Streaming
replication doesn't guarantee that WAL is retained unless you use
replication slots - which you don't appear to be. You can make SR retain
more with approximate methods like wal_keep_segments too, but that's not
a guarantee. From what I can see you're just seeing the combination of
these two limitations, because you don't use the methods to address them
(archive_timeout, replication slots and/or wal_keep_segments).

Greetings,

Andres Freund

--
Sent via pgsql-general mailing list (pgsql-general@postgresql.org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-general