Subject FB 2.0.1 SS - INET errors and gds software consistency checks under Linux
Author Stan
Hi All,

Firebird 2.0.1 SS (debug binary from
Gentoo Linux:
Linux stan-gentoo #2 SMP Fri Sep 14 09:22:10 PDT 2007 i686
Intel(R) Pentium(R) 4 CPU 2.80GHz GenuineIntel GNU/Linux
512 MB Ram
1 IDE disk with plenty of free space

Firebird Log:
stan-gentoo (Server) Fri Sep 21 06:58:37 2007
INET/inet_error: select in packet_receive errno = 9

(Lots of these, ~1 every second)

stan-gentoo (Server) Fri Sep 21 07:00:42 2007
INET/inet_error: read errno = 9

(Lots of these, ~1 every second)

stan-gentoo (Server) Fri Sep 21 07:01:35 2007
Database: data
internal gds software consistency check (partner index
description not found (175), file: idx.cpp line: 1274)
(Once in a while, ~ 1-2 every couple of hours)

stan-gentoo (Client) Fri Sep 21 04:22:03 2007
/opt/firebird/bin/fbguard: bin/fbserver terminated abnormally (-1)
(Once in a while, ~ 1-2 every couple of hours)

These errors are reproducible. I create a new database,
and leave my test overnight and every time I will get the same
errors in the log. This is reproducible on 2 different machines,
the other running Fedora Core 7.

My "test" is an application the uses the C API (ibase.h). My
connections get closed, and I have 0 - 5 connections at one time.

gstat version LI-V2.0.1.12855 Firebird 2.0

Database "/home/stan/data/data.fdb"
Database header page information:
Flags 0
Checksum 12345
Generation 135385
Page size 16384
ODS version 11.0
Oldest transaction 129565
Oldest active 135381
Oldest snapshot 135381
Next transaction 135382
Bumped transaction 1
Sequence number 0
Next attachment ID 0
Implementation ID 19
Shadow count 0
Page buffers 0
Next header page 0
Database dialect 3
Creation date Sep 20, 2007 12:22:27
Attributes force write

Variable header data:

I am using 'cooperative' garbage collection, but I tried 'combined' as
well, results are the same. Default sweep interval (20000).

I am using a matching

I need to know what causes the socket errors and what causes
the gds checks and the crashes.

This is not a "random" physical database corruption since I can
reproduce it and it happens on multiple machines.

thanks in advance,