Code review comment for lp:~vlad-lesin/percona-server/5.6-sql_timeout_twitter
Revision history for this message
| Vlad Lesin (vlad-lesin) wrote | # |
© 2004
Canonical Ltd.
•
Terms of use
•
Data privacy
•
Contact Launchpad Support
•
Blog
•
Careers
•
System status
•
f33e652
(Get the code!)
> I see you encountered same issue as I did also. performance_schema *digest*
> tables are rely heavily on sql_yacc.yy grammar. So if someone just add new
> token (or even rule?) many of statement's digests will change.
...
> Do you have any thoughts on how to workaround this?
Digest calculation is based on token IDs. So every time we add new token in the middle of tokens list we just shift token IDs that are after just inserted token. So I see two variants how to avoid that. The first is add new tokens at the end of the list. The second - calculate digest using text representation of query instead of set of tokens.