Intermittently dbvctl log transfer fails with ERROR_CODE:5 Error executing SQLite statement: database is locked
Problem Details
Problem Statement
Intermittently dbvctl log transfer fails with ERROR_CODE:5 Error executing SQLite statement: database is locked
Applicable Error Code
ERROR_CODE:5 Error executing sqlite statement:database is locked
Affected Versions
v8 and v9
Affected Platforms
All
Description
This error occurs when you have multiple DDC configuration setup on the server and that use the same repository for storing data Ex: /usr/dbvisit/standby/conf/DEV.db. Sometimes processes of both configurations try to access/modify the repository at once and fail with SQLite database lock. We could resolve this issue by increasing the timeout value of internal parameter SQLITE_BUSY_TIMEOUT.
Default value
SQLITE_BUSY_TIMEOUT=5000
Step1. Edit Dbvisit configuration file on your primary server and add this parameter
SQLITE_BUSY_TIMEOUT=20000
Step2: Run the command from primary to copy the updated DDC to standby
./dbvctl -d <DDC> -c
Step3: Restart the daemon process from both primary and standby for changes to take into effect
./dbvctl -d DEV -D stop
./dbvctl -d DEV -D start
A client of ours changed the SQLITE_BUSY_TIMEOUT parameter to 20000 but still these messages still occur. Any other options to resolve this?
Regards.
Hi there, the above issue is fixed in Dbvisit 9.0.16.
DS9-636: Significantly reduced chances of Dbvisit Standby sqlite database locking under heavy workloads.
You could upgrade to latest Dbvisit (9.0.18) and please let us know if this helped
You can download using the below link:
https://dbvisit.com/download-upgrade/dbvisit-standby
Release Notes:
https://dbvisit.atlassian.net/wiki/spaces/DS9QSG/pages/1244987429/Release+Notes#ReleaseNotes-
Upgrade Instructions:
https://dbvisit.atlassian.net/wiki/spaces/DS9QSG/pages/1244987781/Upgrade+Dbvisit+Standby