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

Posted to issues@ratis.apache.org by "Tsz-wo Sze (Jira)" <ji...@apache.org> on 2021/04/20 14:50:00 UTC

[jira] [Updated] (RATIS-829) Ensure eviction of stateMachineData from cache when both followers catch up

     [ https://issues.apache.org/jira/browse/RATIS-829?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Tsz-wo Sze updated RATIS-829:
-----------------------------
    Fix Version/s:     (was: 2.0.0)

Ratis 2.0.0 was already released.  Removing it from "Fix Version/s" here.

> Ensure eviction of stateMachineData from cache when both followers catch up
> ---------------------------------------------------------------------------
>
>                 Key: RATIS-829
>                 URL: https://issues.apache.org/jira/browse/RATIS-829
>             Project: Ratis
>          Issue Type: Bug
>            Reporter: Shashikant Banerjee
>            Assignee: Shashikant Banerjee
>            Priority: Major
>
> Currently, the data in the StateMachineCache is evicted as soon as the applyTransaction call is issued for a transaction in Ratis in ozone. In our testing with keys in few kbs of size, it was figured that the data is evicted from the cache before append requests can be processed in a slightly slow follower thereby making leader read the chunk data from underlying fs/disk very frequently. This leads to slowing down the leader as well as well as overall throughput of the pipeline. 
> The idea here is to ensure the data is evicted from the cache only when both followers have caught up with the match index and for that, it will call into stateMachine if the data is cached in the stateMachine. 
> This is required for https://issues.apache.org/jira/browse/HDDS-3227.



--
This message was sent by Atlassian Jira
(v8.3.4#803005)