[
https://issues.apache.org/jira/browse/STORM-856?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14577958#comment-14577958
]
Jungtaek Lim edited comment on STORM-856 at 6/8/15 10:42 PM:
-------------------------------------------------------------
Yes, I see your point.
Neither thrift and clojure StormBase don't have :delay-secs, so update-storm!
and storm-base doesn't work on :delay-secs.
So it's just added to map on demand in runtime, and when nimbus is restarted,
nimbus doesn't recognize this value.
{code}
(update-storm!
[this storm-id new-elems]
(let [base (storm-base this storm-id nil)
executors (:component->executors base)
new-elems (update new-elems :component->executors (partial merge
executors))]
(set-data cluster-state (storm-path storm-id)
(-> base
(merge new-elems)
thriftify-storm-base
Utils/serialize)
acls)))
(storm-base
[this storm-id callback]
(when callback
(swap! storm-base-callback assoc storm-id callback))
(clojurify-storm-base (maybe-deserialize (get-data cluster-state
(storm-path storm-id) (not-nil? callback)) StormBase)))
{code}
was (Author: kabhwan):
Yes, I see your point.
Neither thrift and clojure StormBase don't have :delay-secs, so update-storm!
doesn't work on :delay-secs.
So it's just added to map on demand in runtime, and when nimbus is restarted,
nimbus doesn't recognize this value.
> If nimbus goes down with a delayed kill topo cmd, it never comes up
> -------------------------------------------------------------------
>
> Key: STORM-856
> URL: https://issues.apache.org/jira/browse/STORM-856
> Project: Apache Storm
> Issue Type: Bug
> Affects Versions: 0.10.0
> Reporter: Derek Dagit
>
> 1. Describe observed behavior.
> nimbus is in a crash loop
> 2. What is the expected behavior?
> nimbus stays up
> 3. Outline the steps to reproduce the problem.
> launch a topo
> kill topo with `storm kill $topo` (no wait argument, so it defaults to 30s)
> Immediately restart nimbus
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
