You are viewing a plain text version of this content. The canonical link for it is here.

Posted to issues@hbase.apache.org by "stack (JIRA)" <ji...@apache.org> on 2018/03/19 21:17:00 UTC

[jira] [Comment Edited] (HBASE-20179) [TESTING] Scale

    [ https://issues.apache.org/jira/browse/HBASE-20179?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16405464#comment-16405464 ] 

stack edited comment on HBASE-20179 at 3/19/18 9:16 PM:
--------------------------------------------------------

Just had a successful 25B ITBLL run w/ monkeys on 7 node cluster. Took about just over a day. Made ~100 regions per server. 22hours to generate and 5 hours to verify. Needs to go faster so can do more cycles! (Don't recall doing 25B on branch-1... maybe someone else did).


was (Author: stack):
Just had a successful 25B ITBLL run w/ monkeys on 7 node cluster. Took about two days. Need to make it go faster so can do more test cycles.

> [TESTING] Scale
> ---------------
>
>                 Key: HBASE-20179
>                 URL: https://issues.apache.org/jira/browse/HBASE-20179
>             Project: HBase
>          Issue Type: Umbrella
>          Components: test
>            Reporter: stack
>            Priority: Critical
>             Fix For: 2.0.0
>
>
> Umbrella issue for scale testing for 2.0.
> At least keep account of what testing has been done in here.
> TODO as subtasks
>  * Long-running test
>  * ITBLL w/o killing Master
>  * Big ITBLL (1B works, 10B needs more verification, 100B, and 1T).
>  * Many Regions (43k regions takes about 12 minutes on a 6node cluster to deploy and 3 1/2 minutes to go down -- needs tuning).



--
This message was sent by Atlassian JIRA
(v7.6.3#76005)