hadoop-hdfs-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "ASF GitHub Bot (JIRA)" <j...@apache.org>
Subject [jira] [Work logged] (HDDS-1624) Refactor operations inside the bucket lock in OM key write.
Date Tue, 04 Jun 2019 15:38:11 GMT

     [ https://issues.apache.org/jira/browse/HDDS-1624?focusedWorklogId=253818&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-253818

ASF GitHub Bot logged work on HDDS-1624:

                Author: ASF GitHub Bot
            Created on: 04/Jun/19 15:37
            Start Date: 04/Jun/19 15:37
    Worklog Time Spent: 10m 
      Work Description: bharatviswa504 commented on pull request #882: HDDS-1624 : Refactor
operations inside the bucket lock in OM key write.
URL: https://github.com/apache/hadoop/pull/882#discussion_r290362757

 File path: hadoop-hdds/common/src/main/java/org/apache/hadoop/utils/db/TypedTable.java
 @@ -79,6 +79,12 @@ public boolean isEmpty() throws IOException {
     return rawTable.isEmpty();
+  @Override
+  public boolean isExist(KEY key) throws IOException {
+    return cache.get(new CacheKey<>(key)) != null ||
 Review comment:
   yes, we fall back. Think of a case like where flush to om DB has not happened, it is only
in the cache. so for key exist we need to check the value also in the cache, as for delete
case we set value as Optional.absent().
   Can you add a java doc to the method caller should acquire/synchronize before calling this
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:

Issue Time Tracking

    Worklog Id:     (was: 253818)
    Time Spent: 2h 50m  (was: 2h 40m)

> Refactor operations inside the bucket lock in OM key write.
> -----------------------------------------------------------
>                 Key: HDDS-1624
>                 URL: https://issues.apache.org/jira/browse/HDDS-1624
>             Project: Hadoop Distributed Data Store
>          Issue Type: Bug
>          Components: Ozone Manager
>            Reporter: Aravindan Vijayan
>            Assignee: Aravindan Vijayan
>            Priority: Critical
>              Labels: pull-request-available
>             Fix For: 0.5.0
>          Time Spent: 2h 50m
>  Remaining Estimate: 0h
> There are a few steps that are done inside the OM bucket lock that are lock invariant
and can be done outside the lock. This patch refactors those steps. It also adds an isExist
API in the metadata store so that we dont need to deserialize the byte[] to Object while doing
a simple _table.get(key) != null_ check. 
> On applying the patch, the OM + SCM (With dummy datanodes) write performance improves
by around 3 - 6x based on number of existing keys in the OM RocksDB. 
> Thanks to [~nandakumar131] who helped with this patch. 

This message was sent by Atlassian JIRA

To unsubscribe, e-mail: hdfs-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: hdfs-issues-help@hadoop.apache.org

View raw message