[ https://issues.apache.org/jira/browse/YARN-6125?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15851674#comment-15851674 ]
Jason Lowe commented on YARN-6125: ---------------------------------- I suppose we could do ellipses, but I'm still a bit confused. Are we both truncating individual messages as they come in as well as the entire diagnostics? I was under the impression we're not worrying about messages but the diagnostics buffer as a whole. Whether they try to fill it with one big message, a few messages, or one character at a time over a zillion calls, I would expect it to look something like this, using the example 1 above as the original diagnostics from the app: {noformat} ...rter than the previour one message3 this is even shorter message4 the shortest one {noformat} Again I think it could be useful to convey to the user how much of the message was truncated (i.e.: showing last N of X bytes) so they have an idea of how much they're missing. But the first cut could just be ellipses at the start and the last N bytes of the diagnostic messages where N is the configured diagnostic limit. > The application attempt's diagnostic message should have a maximum size > ----------------------------------------------------------------------- > > Key: YARN-6125 > URL: https://issues.apache.org/jira/browse/YARN-6125 > Project: Hadoop YARN > Issue Type: Improvement > Components: resourcemanager > Affects Versions: 2.7.0 > Reporter: Daniel Templeton > Assignee: Andras Piros > Priority: Critical > Fix For: 3.0.0-alpha3 > > Attachments: YARN-6125.000.patch, YARN-6125.001.patch, > YARN-6125.002.patch, YARN-6125.003.patch > > > We've found through experience that the diagnostic message can grow > unbounded. I've seen attempts that have diagnostic messages over 1MB. Since > the message is stored in the state store, it's a bad idea to allow the > message to grow unbounded. Instead, there should be a property that sets a > maximum size on the message. > I suspect that some of the ZK state store issues we've seen in the past were > due to the size of the diagnostic messages and not to the size of the > classpath, as is the current prevailing opinion. > An open question is how best to prune the message once it grows too large. > Should we > # truncate the tail, > # truncate the head, > # truncate the middle, > # add another property to make the behavior selectable, or > # none of the above? -- This message was sent by Atlassian JIRA (v6.3.15#6346) --------------------------------------------------------------------- To unsubscribe, e-mail: yarn-issues-unsubscr...@hadoop.apache.org For additional commands, e-mail: yarn-issues-h...@hadoop.apache.org