[jira] [Closed] (GEODE-8076) simplify redis concurrency code

["Nabarun Nag (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/GEODE-8076?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Nabarun Nag closed GEODE-8076.
------------------------------

> simplify redis concurrency code
> -------------------------------
>
>                 Key: GEODE-8076
>                 URL: https://issues.apache.org/jira/browse/GEODE-8076
>             Project: Geode
>          Issue Type: Improvement
>          Components: redis
>            Reporter: Darrel Schneider
>            Assignee: Darrel Schneider
>            Priority: Major
>             Fix For: 1.14.0
>
>
> Currently when doing a redis set operation, for example sadd, the code has to be careful to deal with other threads concurrently changing the same set.
> It does this in a number of ways but this could be simplified by having a higher level layer of the code ensure that for a given redis "key" operations will done in sequential order.
> This can be done safely in a distributed cluster because we now route all operations for a given key to the server that is storing the primary copy of data for that key.
> I spike was done and we found that this form of locking did not hurt performance. Since it allows simpler code that is less likely to have subtle concurrency issues we plan on merging the work done in the spike into the product.



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