Duo Zhang commented on HBASE-15536:

At RS, we will fail the request if the wall edit is too large.

And a simple split does not work, it we fail at the middle, we have no 
information to know if there are more pieces for this edit so either we loss 
data or can not guarantee atomic.
We need to introduce a chunk senmatic for wal edit if we want to do things like 
this. This will also have a benefit that a larger write will not block other 
write operations a lot.

Let me dig more. 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
>            Priority: Critical
>             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

Reply via email to