[ https://issues.apache.org/jira/browse/IGNITE-3621?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15493932#comment-15493932 ]
Andrey Martianov edited comment on IGNITE-3621 at 9/15/16 4:59 PM: ------------------------------------------------------------------- [~sboikov], Thanks! [~amashenkov], Entries expiration is checked on every cache action (in way as point #1 of description says) and processed in threads performed those actions. So I suppose we don't need additional cleaner threads. CleanupWorker really helps only in case if it isn't any cache activity for some time. And I think in this case it will be enough. was (Author: amartianov): [~sboikov], Thanks! [~amashenkov], Entries expiration is checked on every cache action (in way as point #1 of description says) and processed in threads performed that actions. So I suppose we don't need additional cleaner threads. CleanupWorker really helps only in case if it isn't any cache activity for some time. And I think in this case it will be enough. > Make GridCacheTtlManager singleton > ---------------------------------- > > Key: IGNITE-3621 > URL: https://issues.apache.org/jira/browse/IGNITE-3621 > Project: Ignite > Issue Type: Bug > Components: cache > Affects Versions: 1.6 > Reporter: Eduard Shangareev > Assignee: Andrey Martianov > Labels: performance > > Now every cache has own TTL manager, which creates CleanupWorker = new extra > thread. This can cause to extra hundreds of threads (redundant context > switches = performance penalty). > Also, under IGNITE-3513 every put can enter critical section to notify > worker. Obviously, it is not good from performance point of view. > So, my proposal is next: > 1. Expiration should be done on every cache action (on exit thread which > updates cache should invoke {{expire}}). > 2. TtlManager will exist only in one instance. > 3. CleanupWorker will be the only backup if there is no cache activity. It > will wake up with some period to check for work (500 ms, for example). > Moreover, now we keep on-heap pending entries even if a cache is kept > off-head. At least, this issue needs discussion. -- This message was sent by Atlassian JIRA (v6.3.4#6332)