[
https://issues.apache.org/jira/browse/CASSANDRA-18373?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17708064#comment-17708064
]
Yifan Cai commented on CASSANDRA-18373:
---------------------------------------
SSTable import does not currently throw error while the node is draining. The
command completes successfully. The description "SSTables imports that have no
chance of succeeding due to the node being drained" is not entirely true.
We had a discussion offline. Per [~drohrer] and [~semchyshyn], allowing to
import into a draining node seems to be an undefined behavior. Therefore,
adding the patch to clarify the behavior.
> Node Draining Should Abort All Current SSTables Imports
> -------------------------------------------------------
>
> Key: CASSANDRA-18373
> URL: https://issues.apache.org/jira/browse/CASSANDRA-18373
> Project: Cassandra
> Issue Type: Improvement
> Components: Local/Startup and Shutdown
> Reporter: Yuriy Semchyshyn
> Assignee: Yuriy Semchyshyn
> Priority: Normal
> Time Spent: 10m
> Remaining Estimate: 0h
>
> SS tables imports that have no chance of succeeding due to the node being
> drained should fail fast.
> In case there are any SS tables imports currently active, a draining node
> should notify all of them upon flushing its column family stores.
> Each notified SS tables import should abort as soon as possible by throwing
> an `InterruptedException`.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
