You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I`ve run some heavy query on auxiliary (NON-production) database that is placed on same machine as our production. This query was running too long and I decided to kill it:
isql production_alias
SQL>delete from mon$attachments; quit;
Immediatelly after it we detected that no more new connections to production can be established via TCP. Connections to production via local protocol were avaliable.
The firebird.log does NOT contains any interesting info - only messages with errno=-104.
I could create backtrace of fb_smp_server process when we could not to connect (see attach #1) and after restart linux and restoring common workload of out users (see attach #2).
Unfortunatelly I could NOT make log of fb_lock_print -d /var/db/firebird/production.fdb -a -m -c >fb_lock_print_dead.txt because log file has been created with zero size.
SQL> show version;
ISQL Version: LI-V2.5.2.26448 Firebird 2.5 // SUPERCLASSIC
Server version:
Firebird/linux AMD64 (access method), version "LI-V2.5.2.26448 Firebird 2.5"
on disk structure version 11.2
SQL> show database;
Database: kuntsevomain
Owner: SYSDBA
PAGE_SIZE 16384
Number of DB pages allocated = 1888000
Sweep interval = 0
Forced Writes are OFF
Transaction - oldest = 218095653
Transaction - oldest active = 218307431
Transaction - oldest snapshot = 218307431
Transaction - Next = 218365577
ODS = 11.2
Default Character set: WIN1251
Submitted by: @pavel-zotov
Relate to CORE3993
Attachments:
gdb_fb_smp.zip
I`ve run some heavy query on auxiliary (NON-production) database that is placed on same machine as our production. This query was running too long and I decided to kill it:
isql production_alias
SQL>delete from mon$attachments; quit;
Immediatelly after it we detected that no more new connections to production can be established via TCP. Connections to production via local protocol were avaliable.
The firebird.log does NOT contains any interesting info - only messages with errno=-104.
I could create backtrace of fb_smp_server process when we could not to connect (see attach #1) and after restart linux and restoring common workload of out users (see attach #2).
Unfortunatelly I could NOT make log of fb_lock_print -d /var/db/firebird/production.fdb -a -m -c >fb_lock_print_dead.txt because log file has been created with zero size.
SQL> show version;
ISQL Version: LI-V2.5.2.26448 Firebird 2.5 // SUPERCLASSIC
Server version:
Firebird/linux AMD64 (access method), version "LI-V2.5.2.26448 Firebird 2.5"
on disk structure version 11.2
SQL> show database;
Database: kuntsevomain
Owner: SYSDBA
PAGE_SIZE 16384
Number of DB pages allocated = 1888000
Sweep interval = 0
Forced Writes are OFF
Transaction - oldest = 218095653
Transaction - oldest active = 218307431
Transaction - oldest snapshot = 218307431
Transaction - Next = 218365577
ODS = 11.2
Default Character set: WIN1251
Commits: 102700e a8f8465 7ee2887 3746843 c7454a4 74938ad FirebirdSQL/fbt-repository@211856e FirebirdSQL/fbt-repository@6fa99aa FirebirdSQL/fbt-repository@0a2a1c4 FirebirdSQL/fbt-repository@41b2529
The text was updated successfully, but these errors were encountered: