Hi,

A few days ago, the Firebird 1.0 (WI-V6.2.821) server and guardian
crashed on one of our Win2KPro test servers. The only error
messages we got were that the server had terminated abnormally (in
both the interbase.log and Application log on the server) :

MACHINENAME (Client) Wed May 28 00:42:48 2003
C:\Program Files\Firebird\bin\ibserver.exe: terminated
abnormally (-1)

These errors were surrounded by literally THOUSANDS of inet errors
(resulting in the interbase.log file being about 42MB in size):

MACHINENAME (Client) Wed May 28 00:42:49 2003
INET/inet_error: read errno = 10054

There were, unfortunately, no performance monitor counters set up on
the server process, so we don't have that information. We are also,
currently, unable to do a validation on the database, since a long-
running test is being performed on that server, making exclusive
access impossible.

The system is a high-volume 24x7 database which is used as a buffer
database (so data is written by multiple clients and is periodically
cleared as the data is transferred to the main database)

The header page information of the database:

> Database header page information:
> Flags 0
> Checksum 12345
> Generation 5113259
> Page size 4096
> ODS version 10.0
> Oldest transaction 5113253
> Oldest active 5113254
> Oldest snapshot 5113254
> Next transaction 5113255
> Bumped transaction 1
> Sequence number 0
> Next attachment ID 0
> Implementation ID 16
> Shadow count 0
> Page buffers 4096
> Next header page 0
> Database dialect 3
> Creation date Feb 26, 2003 14:48:05
> Attributes
>
> Variable header data:
> *END*

I would like to know what could have caused the crash, but even more
so how the guardian itself crashed. Is this a known issue?

Best Regards,
Riaan