[jira] [Resolved] (HDDS-7917) EC: ECBlockInputStream should try spare replicas on error

["Stephen O'Donnell (Jira)" <ji...@apache.org>]

     [ https://issues.apache.org/jira/browse/HDDS-7917?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Stephen O'Donnell resolved HDDS-7917.
-------------------------------------
    Fix Version/s: 1.4.0
       Resolution: Fixed

> EC: ECBlockInputStream should try spare replicas on error
> ---------------------------------------------------------
>
>                 Key: HDDS-7917
>                 URL: https://issues.apache.org/jira/browse/HDDS-7917
>             Project: Apache Ozone
>          Issue Type: Sub-task
>            Reporter: Stephen O'Donnell
>            Assignee: Stephen O'Donnell
>            Priority: Major
>              Labels: pull-request-available
>             Fix For: 1.4.0
>
>
> In ECBlockInputStream, it is possible to have multiple locations for a single replicaIndex, if, say one replica is decommissioned but still online and there is also an IN_SERVICE replica.
> Similar for Maintenance - there can be maintenance and in_service replicas, or just over replicated replicas.
> As things stand, if a read to an index fails, it causes a failover to reconstruction read, but there are occassions where it would be possible to read from the spare replica and no need to fall back to reconstruction.
> This PR allows spare replicas to be tried before giving up with an error.



--
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