[
https://issues.apache.org/jira/browse/DERBY-2327?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Knut Anders Hatlen updated DERBY-2327:
--------------------------------------
Attachment: niagara-join.png
8cpu-join.png
2cpu-join.png
Also attaching throughput graphs for an identical configuration, only that it
uses the 1000x10000 primary key joins in the DERBY-1961 test client. This load
is more lock intensive than the single-record select load, and the effect of
the patch is therefore greater. All configurations show improved throughput for
concurrent connections (and some show improvement even in the single-user case).
> Reduce monitor contention in LockSet
> ------------------------------------
>
> Key: DERBY-2327
> URL: https://issues.apache.org/jira/browse/DERBY-2327
> Project: Derby
> Issue Type: Sub-task
> Components: Performance, Services
> Affects Versions: 10.3.0.0
> Reporter: Knut Anders Hatlen
> Assigned To: Knut Anders Hatlen
> Priority: Minor
> Attachments: 2cpu-join.png, 2cpu-select.png, 8cpu-join.png,
> 8cpu-select.png, ConcurrentHashMap.diff, derby-2327-1a.diff,
> derby-2327-1a.stat, derby-2327-2a.diff, derby-2327-2a.stat,
> derby-2327-3a.diff, derby-2327-4a.diff, derby-2327-4a.stat, niagara-join.png,
> niagara-select.png
>
>
> When multiple threads enter the lock manager, only one can access the hash
> table in LockSet. To improve scalability on multiple CPUs, it should be
> possible to have more than one thread accessing the lock table at the same
> time.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
