You are viewing a plain text version of this content. The canonical link for it is here.

Posted to commits@samza.apache.org by "Xinyu Liu (JIRA)" <ji...@apache.org> on 2017/11/20 20:40:00 UTC

[jira] [Created] (SAMZA-1505) Fix CheckpointTool writing only one ssp per task

Xinyu Liu created SAMZA-1505:
--------------------------------

             Summary: Fix CheckpointTool writing only one ssp per task
                 Key: SAMZA-1505
                 URL: https://issues.apache.org/jira/browse/SAMZA-1505
             Project: Samza
          Issue Type: Bug
    Affects Versions: 0.14.0
            Reporter: Xinyu Liu
            Assignee: Xinyu Liu
             Fix For: 0.14.0


Currently when using CheckpointTool to write checkpoints, it only writes a checkpoint of a single ssp per task. By debugging the code, looks like the flatMap() on the checkpoint of Optional tuple(taskname -> Map(ssp -> offset)) merges the results by key taskname. So if we have more than one ssps for a task, it will result in only one ssp left after merging. I add a fix by store the results explicitly in a list and then groupBy() on it, which fixed the problem. I still don't understand why scala will merge the key in flatmap.



--
This message was sent by Atlassian JIRA
(v6.4.14#64029)