Bug #56535 | mysqld tries to load triggers on tables in performance_schema database | ||
---|---|---|---|
Submitted: | 3 Sep 2010 12:12 | Modified: | 3 Sep 2010 16:39 |
Reporter: | Shane Bester (Platinum Quality Contributor) | Email Updates: | |
Status: | Not a Bug | Impact on me: | |
Category: | MySQL Server: Performance Schema | Severity: | S5 (Performance) |
Version: | 5.6.0-m4 | OS: | Any |
Assigned to: | Marc ALFF | CPU Architecture: | Any |
[3 Sep 2010 12:12]
Shane Bester
[3 Sep 2010 12:13]
MySQL Verification Team
evidence
Attachment: bug56535_processmonitor_output.txt (text/plain), 2.26 KiB.
[3 Sep 2010 16:39]
Marc ALFF
It is correct that the server code will attempt to find a trigger for performance schema tables. This is because no specific hard coded logic for the performance schema was added in the trigger implementation. Now, I do not understand why this qualifies as a "performance problem": one the table definition, with no trigger, is loaded in the server, this table definition will stay in the table definition cache and the trigger file will not be re-opened ... If too many file io happens on the .TRG files, it's because the table definition cache is too small that table definitions got evicted, which is a performance tuning problem that affects the server in general, and every table in the application (not just the performance schema). This works as designed. Closing as not a bug.