[
https://issues.apache.org/jira/browse/HBASE-5952?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13269995#comment-13269995
]
Zhihong Yu commented on HBASE-5952:
-----------------------------------
Patch makes sense.
There was a conflict in HTableDescriptor.java:
{code}
***************
*** 164,170 ****
* Constant that denotes the maximum default size of the memstore after
which
* the contents are flushed to the store files
*/
- public static final long DEFAULT_MEMSTORE_FLUSH_SIZE = 1024*1024*64L;
private volatile Boolean meta = null;
private volatile Boolean root = null;
--- 164,170 ----
* Constant that denotes the maximum default size of the memstore after
which
* the contents are flushed to the store files
*/
+ public static final long DEFAULT_MEMSTORE_FLUSH_SIZE = 1024*1024*128L;
private volatile Boolean meta = null;
private volatile Boolean root = null;
{code}
> Sync hardcoded default flush size and max file size with hbase-default.xml
> --------------------------------------------------------------------------
>
> Key: HBASE-5952
> URL: https://issues.apache.org/jira/browse/HBASE-5952
> Project: HBase
> Issue Type: Bug
> Affects Versions: 0.92.2
> Reporter: Jimmy Xiang
> Assignee: Jimmy Xiang
> Priority: Minor
> Fix For: 0.92.2
>
> Attachments: hbase-5952.patch
>
>
> Since the hardcoded default flush size is 64MB, but the default in
> hbase-default.xml is 128MB, if the client does set it to 64MB,
> the actual flush size will be 128MB instead, due to the way HRegion get the
> flush size. We can change how HRegion get the flush size,
> but it is clean and simple to sync up the defaults.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
