[ https://issues.apache.org/jira/browse/CASSANDRA-7392?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14627386#comment-14627386 ]
Stefania commented on CASSANDRA-7392: ------------------------------------- Thank you for reviewing this. bq. Small aside: I'm really not a huge fan of having paths specific to tests in the middle of actually used code (as in ReadCommand). I understand there isn't a much easier way to test this currently and I don't oppose this in the short term, but it would make me very happy if we could find a cleaner way to deal with this in the longer run. Maybe we could have a way to mock up a SSTableReader (and to inject it so it's used by queries), which would allow to make it run very slowly. Or something like that. I don't like this either. I already mentioned something similar [here|https://issues.apache.org/jira/browse/CASSANDRA-9102?focusedCommentId=14617799&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-14617799] as this is not the first time I have to use a system property to test a faulty code path. Perhaps CASSANDRA-9165 should be extended to cover all the cases we need (throwing exceptions, taking longer, etc) or another ticket should be open to have a reliable way to inject test code or faults when running dtests. > Abort in-progress queries that time out > --------------------------------------- > > Key: CASSANDRA-7392 > URL: https://issues.apache.org/jira/browse/CASSANDRA-7392 > Project: Cassandra > Issue Type: New Feature > Components: Core > Reporter: Jonathan Ellis > Assignee: Stefania > Fix For: 3.x > > > Currently we drop queries that time out before we get to them (because node > is overloaded) but not queries that time out while being processed. > (Particularly common for index queries on data that shouldn't be indexed.) > Adding the latter and logging when we have to interrupt one gets us a poor > man's "slow query log" for free. -- This message was sent by Atlassian JIRA (v6.3.4#6332)