[
https://issues.apache.org/jira/browse/DRILL-3103?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Chris Westin updated DRILL-3103:
--------------------------------
Attachment: DRILL-3103.1.patch.txt
Trial patch that bumps the last activity time for writes as well as reads. We
suspect that the large amounts of data being read in parallel (by this simple
query) is just flooding the pipeline back to the client, so it isn't getting a
chance to see the heartbeats that it uses to determine that the connection is
still alive.
https://reviews.apache.org/r/34297/
> EncoderException: RpcEncoder must produce at least one message.
> ---------------------------------------------------------------
>
> Key: DRILL-3103
> URL: https://issues.apache.org/jira/browse/DRILL-3103
> Project: Apache Drill
> Issue Type: Bug
> Components: Execution - RPC
> Reporter: Rahul Challapalli
> Assignee: Chris Westin
> Attachments: DRILL-3103.1.patch.txt, error.log
>
>
> git.commit.id.abbrev=d8b1975
> I ran the below query on TPCH SF100 data on a 2 node cluster. The query
> failed for some reason. I attached the logs
> Query
> {code}
> select o_orderkey from orders;
> {code}
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
