--- In firebird-support@yahoogroups.com, Helen Borrie <helebor@...> wrote:

> 1. In the log, is it preceded by an 'Internal gds inconsistency check (some text or a number)' error?
>
> 2. What version and model of Firebird?
>
> ./heLen
>

Version is 2.11, SS.

I have had quite a lot of trouble with this particular installation, and I am beginning to suspect the hardware.

There is nothing in the log entry immediately preceding the above error, apart from occasional "INET/inet_error: read errno = 10054". However, the same procedure has written the error to the log intermittently over the past few weeks. Last week we had major corruption, which had to be fixed by dropping a corrupt table, backing up and restoring.

Here are some of the messages:

NFPSERVER (Server) Wed Mar 11 15:01:36 2009
Database: NFP
lock conflict on no wait transaction
deadlock
update conflicts with concurrent update
concurrent transaction number is 1701779
internal gds software consistency check (decompression overran buffer (179), file: sqz.cpp line: 231)


NFPSERVER (Server) Wed Mar 11 15:01:36 2009
Database: NFP
internal gds software consistency check (Attempt to call GlobalRWLock::unlock() while not holding a valid lock for logical owner)


NFPSERVER (Server) Wed Mar 11 15:22:20 2009
Database: NFP
Relation has 133 orphan backversions (1732 in use) in table BATCH_EVENT (147)


NFPSERVER (Server) Wed Mar 11 15:22:21 2009
Database: NFP
Index 2 is corrupt on page 84938 level 1. File: ..\..\..\src\jrd\validation.cpp, line: 1658
in table BATCH_EVENT_BED (148)


NFPSERVER (Server) Wed Mar 11 15:22:21 2009
Database: NFP
Page 17200 wrong type (expected 7 encountered 5)

Any clues greatly appreciated.

Roger Vellacott
Passfield Data Systems Ltd