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

Posted to commits@beam.apache.org by "Amit Sela (JIRA)" <ji...@apache.org> on 2016/09/20 18:59:20 UTC

[jira] [Created] (BEAM-648) Persist and restore Aggergator values in case of recovery from failure

Amit Sela created BEAM-648:
------------------------------

             Summary: Persist and restore Aggergator values in case of recovery from failure
                 Key: BEAM-648
                 URL: https://issues.apache.org/jira/browse/BEAM-648
             Project: Beam
          Issue Type: Bug
          Components: runner-spark
            Reporter: Amit Sela
            Assignee: Amit Sela


While Aggregators are fault-tolerant (pending https://github.com/apache/incubator-beam/pull/909), they will recover into their initial value. 

The SparkRunner should persist the values to durable storage and recover from the persisted state.

One way to go would be to persist via Spark's listeners, for example "onBatchCompleted".

For more details see SPARK-5206.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)