SEO

Hi

in a 5.0.45 linux master-master replication setup, I see the error
message from the subject every now and then. Sometimes it does not
happen for a couple of months, then it happens a couple of times a day.
I cannot see any network problems otherwise, and a memtest did not bring
up any memory problems. It "feels" like this happens on a rather high
load on the system, as it never happened at night.

What can I do about that?

Regards
Dominik

Hi.

Juan Eduardo Moreno wrote:
> Sorry, what is your setting for max_allowed_packet variable?

mysql> show variables like "%allowed%";
+--------------------+----------+
| Variable_name | Value |
+--------------------+----------+
| max_allowed_packet | 16776192 |
+--------------------+----------+
1 row in set (0.00 sec)

Here's an excerpt from the logs. It does not look like that value is
involved in this:

080428 13:07:04 [ERROR] Error reading packet from server: Out of memory
(Needed 2704 bytes) ( server_errno=5)
080428 13:07:04 [Note] Slave I/O thread: Failed reading log event,
reconnecting to retry, log 'SRV02bin.000131' position 447892967
080428 13:07:04 [Note] Slave: connected to master
'mmrepli@10.250.250.32:3306',replication resumed in log
'SRV02-bin.000131' at position 447892967
080428 13:22:14 [ERROR] Error reading packet from server: Out of memory
(Needed 2704 bytes) ( server_errno=5)
080428 13:22:14 [Note] Slave I/O thread: Failed reading log event,
reconnecting to retry, log 'SRV02-bin.000131' position 471157588
080428 13:22:14 [Note] Slave: connected to master
'mmrepli@10.250.250.32:3306',replication resumed in log
'SRV02-bin.000131' at position 471157588
080428 14:38:06 [ERROR] Error reading packet from server: Out of memory
(Needed 3464 bytes) ( server_errno=5)
080428 14:38:06 [Note] Slave I/O thread: Failed reading log event,
reconnecting to retry, log 'SRV02-bin.000131' position 583864714
080428 14:38:06 [Note] Slave: connected to master
'mmrepli@10.250.250.32:3306',replication resumed in log
'SRV02-bin.000131' at position 583864714
080428 15:13:24 [ERROR] Error reading packet from server: Out of memory
(Needed 3128 bytes) ( server_errno=5)
080428 15:13:24 [Note] Slave I/O thread: Failed reading log event,
reconnecting to retry, log 'SRV02-bin.000131' position 635409380
080428 15:13:24 [Note] Slave: connected to master
'mmrepli@10.250.250.32:3306',replication resumed in log
'SRV02-bin.000131' at position 635409380
080428 15:32:14 [ERROR] Error reading packet from server: Out of memory
(Needed 2688 bytes) ( server_errno=5)
080428 15:32:14 [Note] Slave I/O thread: Failed reading log event,
reconnecting to retry, log 'SRV02-bin.000131' position 665146758
080428 15:32:14 [Note] Slave: connected to master
'mmrepli@10.250.250.32:3306',replication resumed in log
'SRV02-bin.000131' at position 665146758

Regards
Dominik

Hi,
This error occur when slave servers could incorrectly interpret an
out-of-memory error from the master and reconnect using the wrong binary log
position.

This was fix in 5.0.48 version. Please, try to update your version ( from
5.0.45) of mysql and try again.

regards,
Juan


On 4/28/08, Dominik Klein <dk@in-telegence.net> wrote:
>
> Hi.
>
> Juan Eduardo Moreno wrote:
>
> > Sorry, what is your setting for max_allowed_packet variable?
> >
>
> mysql> show variables like "%allowed%";
> +--------------------+----------+
> | Variable_name | Value |
> +--------------------+----------+
> | max_allowed_packet | 16776192 |
> +--------------------+----------+
> 1 row in set (0.00 sec)
>
> Here's an excerpt from the logs. It does not look like that value is
> involved in this:
>
> 080428 13:07:04 [ERROR] Error reading packet from server: Out of memory
> (Needed 2704 bytes) ( server_errno=5)
> 080428 13:07:04 [Note] Slave I/O thread: Failed reading log event,
> reconnecting to retry, log 'SRV02bin.000131' position 447892967
> 080428 13:07:04 [Note] Slave: connected to master '
> mmrepli@10.250.250.32:3306',replication resumed in log 'SRV02-bin.000131'
> at position 447892967
> 080428 13:22:14 [ERROR] Error reading packet from server: Out of memory
> (Needed 2704 bytes) ( server_errno=5)
> 080428 13:22:14 [Note] Slave I/O thread: Failed reading log event,
> reconnecting to retry, log 'SRV02-bin.000131' position 471157588
> 080428 13:22:14 [Note] Slave: connected to master '
> mmrepli@10.250.250.32:3306',replication resumed in log 'SRV02-bin.000131'
> at position 471157588
> 080428 14:38:06 [ERROR] Error reading packet from server: Out of memory
> (Needed 3464 bytes) ( server_errno=5)
> 080428 14:38:06 [Note] Slave I/O thread: Failed reading log event,
> reconnecting to retry, log 'SRV02-bin.000131' position 583864714
> 080428 14:38:06 [Note] Slave: connected to master '
> mmrepli@10.250.250.32:3306',replication resumed in log 'SRV02-bin.000131'
> at position 583864714
> 080428 15:13:24 [ERROR] Error reading packet from server: Out of memory
> (Needed 3128 bytes) ( server_errno=5)
> 080428 15:13:24 [Note] Slave I/O thread: Failed reading log event,
> reconnecting to retry, log 'SRV02-bin.000131' position 635409380
> 080428 15:13:24 [Note] Slave: connected to master '
> mmrepli@10.250.250.32:3306',replication resumed in log 'SRV02-bin.000131'
> at position 635409380
> 080428 15:32:14 [ERROR] Error reading packet from server: Out of memory
> (Needed 2688 bytes) ( server_errno=5)
> 080428 15:32:14 [Note] Slave I/O thread: Failed reading log event,
> reconnecting to retry, log 'SRV02-bin.000131' position 665146758
> 080428 15:32:14 [Note] Slave: connected to master '
> mmrepli@10.250.250.32:3306',replication resumed in log 'SRV02-bin.000131'
> at position 665146758
>
> Regards
> Dominik
>
> --
> MySQL General Mailing List
> For list archives: http://lists.mysql.com/mysql
> To unsubscribe:
> http://lists.mysql.com/mysql?unsub=j...domc@gmail.com
>
>

Upgrade to 5.0.51b seems to have solved the problem. On wednesday, I saw
the problem about a dozen times during a peak time. Upgraded wednesday
night and have not seen the error since.

Thanks
Dominik

Juan Eduardo Moreno wrote:
> Hi,
> This error occur when slave servers could incorrectly interpret an
> out-of-memory error from the master and reconnect using the wrong binary
> log position.
>
> This was fix in 5.0.48 version. Please, try to update your version (
> from 5.0.45) of mysql and try again.
>
> regards,
> Juan