maedhroz commented on code in PR #2935:
URL: https://github.com/apache/cassandra/pull/2935#discussion_r1442402238
##########
src/java/org/apache/cassandra/index/sai/StorageAttachedIndex.java:
##########
@@ -493,8 +493,7 @@ public long getEstimatedResultRows()
@Override
public boolean isQueryable(Status status)
{
- // consider unknown status as queryable, because gossip may not be
up-to-date for newly joining nodes.
- return status == Status.BUILD_SUCCEEDED || status == Status.UNKNOWN;
+ return status == Status.BUILD_SUCCEEDED;
Review Comment:
The main reason I changed this is that it made my more lightweight mechanism
for avoiding the middle node in the test matrix work with less effort. However,
I'm not entirely sure why we've ever considered UNKNOWN to be a queryable
state. If gossip hasn't propagated yet, was the idea to just fire queries at
the new node and hope the index is queryable locally? Is doing the simpler
thing (only using BUILD_SUCCEEDED) going to cause a major availability problem?
CC @mike-tr-adamson @adelapena
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
