Future reference:
Problem was caused by primary DNS resolver suffering from degraded
performance. Server experienced a software issue with BIND, the
software that performs DNS resolution.
On Wed, May 18, 2011 at 10:11 AM, Sam Stearns <samtstearns@gmail.com> wrote:
> Thanks, Tim.
>
> I've turned up logging some more.
>
> On Wed, May 18, 2011 at 8:38 AM, Tim <elatllat@gmail.com> wrote:
>> First google hit said this:
>>
>> "It seems (not sure) to occur when user retypes password fast.
>> Maybe in this case ODBC driver gets old connection somehow from driver
>> manager."
>>
>> also
>> "It looks to me like the ODBC driver is failing to follow the protocol.
>> It shouldn't respond with an 'X' if it doesn't have a password, just
>> disconnect."
>> I would start logging traffic so you have more to go on next time it
>> happens.
>>
>> On Tue, May 17, 2011 at 6:52 PM, Sam Stearns <samtstearns@gmail.com> wrote:
>>>
>>> Anyone have any ideas on this? We're an ISP. I was thinking of the
>>> possibility of a hack attempt.
>>>
>>> On Tue, May 17, 2011 at 6:52 PM, Sam Stearns <samtstearns@gmail.com>
>>> wrote:
>>> > Howdy,
>>> >
>>> > Environment:
>>> >
>>> > Postgres 8.2.3
>>> > Solaris 10
>>> >
>>> > We were unable to make a remote postgres connection for ~2 hours this
>>> > afternoon. Receiving the following message in the postgres log every
>>> > 30 seconds:
>>> >
>>> > LOG: expected password response, got message type 88
>>> >
>>> > Connection limit is 300. 17 active connections at the time. Bounced
>>> > postgres successfully but the above error message was still being
>>> > logged. About 20 minutes later the error message stopped and
>>> > connections started to re-establish.
>>> >
>>> > Any ideas?
>>> >
>>> > Thank you,
>>> >
>>> > Sam
>>> >
>>>
>>> --
>>> Sent via pgsql-admin mailing list (pgsql-admin@postgresql.org)
>>> To make changes to your subscription:
>>> http://www.postgresql.org/mailpref/pgsql-admin
>>
>>
>