[
https://issues.apache.org/jira/browse/HBASE-30162?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18082710#comment-18082710
]
Prathyusha commented on HBASE-30162:
------------------------------------
_This appears to be a duplicate of HBASE-27091 ("Speed up the loading of table
descriptor from filesystem"), which already added parallel_ _.tableinfo_
_loading to FSTableDescriptors.getAll(), will close this one_
> Parallelise .tableinfo file reads in FSTableDescriptors.getAll() during
> finishActiveMasterInitialization
> --------------------------------------------------------------------------------------------------------
>
> Key: HBASE-30162
> URL: https://issues.apache.org/jira/browse/HBASE-30162
> Project: HBase
> Issue Type: Improvement
> Reporter: Prathyusha
> Assignee: Prathyusha
> Priority: Minor
>
> Sometimes when the underlying FileSystem is facing any partial AZ outages or
> the DN layer is too slow, the sequential read of tableinfo files slows down
> the Master initialisation, the plan is to use a threadpool to read tableinfo
> files of different tables in parallel.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
