You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@hbase.apache.org by "Jean-Daniel Cryans (JIRA)" <ji...@apache.org> on 2010/02/07 06:50:28 UTC
[jira] Created: (HBASE-2188) Group ICVs waiting on row lock
Group ICVs waiting on row lock
------------------------------
Key: HBASE-2188
URL: https://issues.apache.org/jira/browse/HBASE-2188
Project: Hadoop HBase
Issue Type: Improvement
Reporter: Jean-Daniel Cryans
Fix For: 0.21.0
In the same spirit as group commit down in HLog, incrementColumnValue should be grouping all requests waiting on the row lock into one in HRegion. This means that 10 clients waiting on the lock should be all summed up and then applied only once.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Updated: (HBASE-2188) Group ICVs waiting on row lock
Posted by "Jean-Daniel Cryans (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-2188?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Jean-Daniel Cryans updated HBASE-2188:
--------------------------------------
Priority: Minor (was: Major)
Fix Version/s: (was: 0.21.0)
Setting this to minor priority and not for 0.21, let's revisit if we need to.
> Group ICVs waiting on row lock
> ------------------------------
>
> Key: HBASE-2188
> URL: https://issues.apache.org/jira/browse/HBASE-2188
> Project: Hadoop HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Priority: Minor
>
> In the same spirit as group commit down in HLog, incrementColumnValue should be grouping all requests waiting on the row lock into one in HRegion. This means that 10 clients waiting on the lock should be all summed up and then applied only once.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.
[jira] Commented: (HBASE-2188) Group ICVs waiting on row lock
Posted by "ryan rawson (JIRA)" <ji...@apache.org>.
[ https://issues.apache.org/jira/browse/HBASE-2188?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12830693#action_12830693 ]
ryan rawson commented on HBASE-2188:
------------------------------------
for what it's worth, my previous testing in 0.20 indicated a single cell can take upwards of 2b ICV calls/day. The actual contention is fairly low. We should weigh the implementation complexity against using a sharded counter (which you'd need for multi-DC counters anyways) to achieve truly extreme counter performance.
> Group ICVs waiting on row lock
> ------------------------------
>
> Key: HBASE-2188
> URL: https://issues.apache.org/jira/browse/HBASE-2188
> Project: Hadoop HBase
> Issue Type: Improvement
> Reporter: Jean-Daniel Cryans
> Fix For: 0.21.0
>
>
> In the same spirit as group commit down in HLog, incrementColumnValue should be grouping all requests waiting on the row lock into one in HRegion. This means that 10 clients waiting on the lock should be all summed up and then applied only once.
--
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.