[
https://issues.apache.org/jira/browse/IGNITE-4552?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Semen Boikov reassigned IGNITE-4552:
------------------------------------
Assignee: Stanilovsky Evgeny (was: Semen Boikov)
Reviewed changes, my comments:
- it does not compile with jdk7, please switch to jdk7
- 'while (rmvQueue.peekFirst().expireTime() < U.currentTimeMillis())' - you
need check result of 'peekFirst()' for null
- do not schedule cleanup task per-partition, it should be single task per node
- error in 'onDeferredDelete', now if size > limit you remove all itmes from
queue, actually you need pollFirst while size > limit
- queue.peekFirst.expireTime > 'current time' is necessary in
'onDeferredDelete' since there can be too many expired entries and cleanup task
will not be able to do cleanup fast enough
- please follow coding conventions (remove unused imports, add javadoc for all
methods, fields, 'Override' must be on the same line with method)
Thnaks
> Optimize GridDhtLocalPartition.rmvQueue
> ---------------------------------------
>
> Key: IGNITE-4552
> URL: https://issues.apache.org/jira/browse/IGNITE-4552
> Project: Ignite
> Issue Type: Bug
> Components: cache
> Affects Versions: 1.6
> Reporter: Alexei Scherbakov
> Assignee: Stanilovsky Evgeny
> Fix For: 2.0
>
> Attachments: benchmark-put-remove-simultaneously.properties,
> Plot_ThroughputLatencyProbe_01_31_origin.png,
> Plot_ThroughputLatencyProbe_01_4552_mine.png,
> Plot_ThroughputLatencyProbe_01_mine_3node.png,
> Plot_ThroughputLatencyProbe_01_mine.png,
> Plot_ThroughputLatencyProbe_01_origin_3node.png,
> Plot_ThroughputLatencyProbe_01_origin.png, Screenshot_20170124_155355.png
>
>
> Current implementation stores deferred entry removals in rmvQueue for
> consistency guaranties.
> This can lead to significant heap over-usage(I observed several Gbs) in case
> of many caches with removals, because currently queue is cleared lazily after
> reaching max capacity(200_000 by default).
> This can be mitigated by using lower IGNITE_ATOMIC_CACHE_DELETE_HISTORY_SIZE,
> but can lead to consistency issues in case of frequent cache updates.
> Possible optimizations:
> * Use single fixed size queue per all caches to overcome limitations of
> IGNITE_ATOMIC_CACHE_DELETE_HISTORY_SIZE workaround.
> * Do queue cleaning in background
> * Move queue to an off-heap.
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
