[jira] [Commented] (HDDS-7601) [Snapshot] Use RocksDB to persist compaction log

["Hemant Kumar (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/HDDS-7601?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17766618#comment-17766618 ] 

Hemant Kumar commented on HDDS-7601:
------------------------------------

As part of PR: [5303|https://github.com/apache/ozone/pull/5303], a new column family is added to RocksDB.

A follow up jira: https://issues.apache.org/jira/browse/HDDS-9311

> [Snapshot] Use RocksDB to persist compaction log
> ------------------------------------------------
>
>                 Key: HDDS-7601
>                 URL: https://issues.apache.org/jira/browse/HDDS-7601
>             Project: Apache Ozone
>          Issue Type: Sub-task
>            Reporter: Siyao Meng
>            Assignee: Hemant Kumar
>            Priority: Minor
>              Labels: pull-request-available
>
> 1. We need to deal with potential (silent) corruption against the compaction log. We could leverage RocksDB's built-in [per-key checksum|https://rocksdb.org/blog/2022/07/18/per-key-value-checksum.html] to deal with it rather than handling it ourselves.
> 2. Implement an {{OmDBMetaInfoManager}} interface that could use either plain text or RocksDB as the backend storage. Or maybe just RocksDB for simplicity.
> 3. This means each OM would have its own metadata DB that stores information about the main OM DB, where we could store compaction log, SST filtering status and potentially more.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@ozone.apache.org
For additional commands, e-mail: issues-help@ozone.apache.org