Lars Volker has posted comments on this change. Change subject: IMPALA-4494: Fix crash in SimpleScheduler ......................................................................
Patch Set 1: Thanks for your feedback. I agree that this is not optimal, and if there was a way to send the equivalent of EAGAIN to the client and make it try again either with this or a different coordinator that could be a workaround. I'd also be happy to add a small timeout of a few seconds before returning an error. I'm rather reluctant to change the way we register the local backend with itself, since that way we will have two mechanisms of updating the membership information and need to handle conflicts as well. That being said, between a) Relying on the client to try again, b) waiting for 5 seconds (or whatever value), and c) reworking the cluster membership information management code in the scheduler, which one do you prefer? -- To view, visit http://gerrit.cloudera.org:8080/5127 To unsubscribe, visit http://gerrit.cloudera.org:8080/settings Gerrit-MessageType: comment Gerrit-Change-Id: I6e1196a2fa47e5954c4a190aa326c135d039a77f Gerrit-PatchSet: 1 Gerrit-Project: Impala-ASF Gerrit-Branch: master Gerrit-Owner: Lars Volker <l...@cloudera.com> Gerrit-Reviewer: Lars Volker <l...@cloudera.com> Gerrit-Reviewer: Tim Armstrong <tarmstr...@cloudera.com> Gerrit-HasComments: No