This content has been marked as final.
Show 18 replies
-
15. Re: Can we introduce a new managed connection pool implementation that reduces thread contention under load?
jesper.pedersen Jun 12, 2014 3:30 PM (in response to johara)A quick scan and it looks like we are ready for an in-depth code review for inclusion.
Please, squash your commits, and submit a pull requests against the 1.2 branch referencing a JIRA (Feature Request / Major / Core).
We will take minor corrections on the pull request, and any major stuff is found here. Please, add links to the JIRA and pull request to this thread.
Thanks !
-
17. Re: Re: Can we introduce a new managed connection pool implementation that reduces thread contention under load?
jesper.pedersen Jun 17, 2014 9:15 AM (in response to johara)Initial run of 7c40957b8b9ce69557608bad46e2216cb9100e57
NoLoad: DO_RAMPUP=true
Load: DO_RAMPUP=false, TX_BEGIN_DURATION=2L, TX_COMMIT_DURATION=10L
Average over 3 runs for each.
-
sal-load.dat.zip 279 bytes
-
sal-load.png 15.0 KB
-
sal-noload.dat.zip 433 bytes
-
sal-noload.png 17.5 KB
-
scq-load.dat.zip 280 bytes
-
scq-load.png 15.2 KB
-
scq-noload.dat.zip 442 bytes
-
scq-noload.png 16.8 KB
-
-
18. Re: Can we introduce a new managed connection pool implementation that reduces thread contention under load?
garage_logician Jul 29, 2015 2:32 PM (in response to jesper.pedersen)Has this been implemented? If so which version is it in? I am seeing thread locks under high loads in SemaphoreArrayListManagedConnectionPool.returnConnection().