Github user vanzin commented on a diff in the pull request: https://github.com/apache/spark/pull/7839#discussion_r36354984 --- Diff: network/yarn/src/main/java/org/apache/spark/network/yarn/YarnShuffleService.java --- @@ -100,11 +121,34 @@ private boolean isAuthenticationEnabled() { */ @Override protected void serviceInit(Configuration conf) { + + // In case this NM was killed while there were running spark applications, we need to restore + // lost state for the existing executors. We look for an existing file in the NM's local dirs. + // If we don't find one, then we choose a file to use to save the state next time. However, we + // do *not* immediately register all the executors in that file, just in case the application + // was terminated while the NM was restarting. We wait until yarn tells the service about the + // app again via #initializeApplication, so we know it's still running. That is important + // for preventing a leak where the app data would stick around *forever*. This does leave + // a small race -- if the NM restarts *again*, after only some of the existing apps have been --- End diff -- Sorry if I missed, but I don't remember seeing a reply about why is it ok to introduce this race instead of relying on YARN behaving correctly? e.g. always re-register everything and use `stopApplication` to clean things up. You could also clean up the file in `serviceStop()` (depending on what the semantics for that method are).
--- If your project is set up for it, you can reply to this email and have your reply appear on GitHub as well. If your project does not have this feature enabled and wishes so, or if the feature is enabled but not working, please contact infrastructure at infrastruct...@apache.org or file a JIRA ticket with INFRA. --- --------------------------------------------------------------------- To unsubscribe, e-mail: reviews-unsubscr...@spark.apache.org For additional commands, e-mail: reviews-h...@spark.apache.org