[
https://issues.apache.org/jira/browse/IGNITE-16802?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17523884#comment-17523884
]
Vladislav Pyatkov commented on IGNITE-16802:
--------------------------------------------
LGTM
> Raft log storage creates rocksdb instance per group
> ---------------------------------------------------
>
> Key: IGNITE-16802
> URL: https://issues.apache.org/jira/browse/IGNITE-16802
> Project: Ignite
> Issue Type: Bug
> Reporter: Semyon Danilov
> Assignee: Semyon Danilov
> Priority: Blocker
> Labels: ignite-3
> Fix For: 3.0.0-alpha5
>
> Time Spent: 1h 20m
> Remaining Estimate: 0h
>
> At this moment, every raft group creates RocksDBLogStorage which, in turn,
> creates an instance of RocksDB without sharing resources with other log
> storages. This inflicts memory overhead and leads to out of memory killer
> killing process on TeamCity.
> We need to use one rocksdb instance for all log storages
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
