[
https://issues.apache.org/jira/browse/HBASE-15536?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15540424#comment-15540424
]
Duo Zhang commented on HBASE-15536:
-----------------------------------
Seems there is no key value size limitation at server side currently. And the
max key-value size is only used to limit one key-value, which means you can add
100 cells into one put object and each cell is 1MB, then the size of the put is
100MB and breaks our {{AsyncFSWAL}}...
We can introduce a new config named 'max.wal.edit.size' which is used to limit
the size of one atomic operation and set the default value less than 16MB? This
could make it possible to make AsyncFSWAL as our default WAL.
BTW, I think we could add the 'max.wal.edit.size' as a new stop condition when
grab row locks in the multi operation as multi does not guarantee cross row
atomic. Then AsyncFSWAL could also work with batch operation larger than 16MB
unless the size of one row is larger than 16MB or MultiRowMutationEndpoint is
used.
What do you think? [~stack] [~anoopsamjohn]. Too many configs maybe... Thanks.
> Make AsyncFSWAL as our default WAL
> ----------------------------------
>
> Key: HBASE-15536
> URL: https://issues.apache.org/jira/browse/HBASE-15536
> Project: HBase
> Issue Type: Sub-task
> Components: wal
> Affects Versions: 2.0.0
> Reporter: Duo Zhang
> Assignee: Duo Zhang
> Fix For: 2.0.0
>
> Attachments: HBASE-15536-v1.patch, HBASE-15536-v2.patch,
> HBASE-15536-v3.patch, HBASE-15536-v4.patch, HBASE-15536-v5.patch,
> HBASE-15536.patch
>
>
> As it should be predicated on passing basic cluster ITBLL
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
