hbase-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Jean-Daniel Cryans (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (HBASE-8450) Update hbase-default.xml and general recommendations to better suit current hw, h2, experience, etc.
Date Tue, 07 May 2013 00:22:16 GMT

    [ https://issues.apache.org/jira/browse/HBASE-8450?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13650278#comment-13650278
] 

Jean-Daniel Cryans commented on HBASE-8450:
-------------------------------------------

hbase-default.xml:

 - hfile.block.cache.size=0.4: Since we let users dedicate up to 80% of the heap for the block
cache and Memstores, we should fully use that space by default. I've stopped counting the
numbers I've told people that they had all this memory that's not being used because they
were keeping the configs default... Right now we're at 25/40 so I'd just bump up the block
cache to 40. Even better would be doing HBASE-5349...
 - hbase.hregion.majorcompaction=0: disable major compactions, I can't remember the last time
I told someone to run them more often, but just last week I told someone to either disable
them or have them running weekly.
 - hbase.regionserver.nbreservationblocks=0: do we even want this now that we kill on OOME?

 - hbase.regionserver.handler.count=50: or it could be 100, I don't really have data on this.
 - hbase.client.pause=100: make everything snappier? might want to bump up the number of retries.
I don't have data on this.
 - hbase.rpc.timeout=10000: snappier I said!
 - hbase.hstore.compactionThreshold=5: we now have a better compaction algorithm, maybe it's
worth keeping more files around and be more efficient about the whole process?
 - hbase.hstore.blockingStoreFiles=10: I'd give some room to the algo to make good choices
so bumping that up too.
 - zookeeper.session.timeout=90000: is 3 minutes still a good default? Now that we have MSLAB
it's rare that I see legit GC issues.

Other:

 - Enable bloom filters by default, will also help offsetting the higher number of files.
 - Set the default heap higher? 2GB?
                
> Update hbase-default.xml and general recommendations to better suit current hw, h2, experience,
etc.
> ----------------------------------------------------------------------------------------------------
>
>                 Key: HBASE-8450
>                 URL: https://issues.apache.org/jira/browse/HBASE-8450
>             Project: HBase
>          Issue Type: Task
>          Components: Usability
>            Reporter: stack
>            Priority: Critical
>             Fix For: 0.95.1
>
>         Attachments: 8450.txt
>
>
> This is a critical task we need to do before we release; review our defaults.
> On cursory review, there are configs in hbase-default.xml that no longer have matching
code; there are some that should be changed because we know better now and others that should
be amended because hardware and deploys are bigger than they used to be.
> We could also move stuff around so that the must-edit stuff is near the top (zk quorum
config. is mid-way down the page) and beef up the descriptions -- especially since these descriptions
shine through in the refguide.
> Lastly, I notice that our tgz does not "include" an hbase-default.xml other than the
one bundled up in the jar.  Maybe we should make it more accessible.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message