GitHub user liu-zhaokun opened a pull request:
https://github.com/apache/storm/pull/2365
[STORM-2773]If a drpcserver node in cluster is down,drpc cluster won't work
if we don't modify the drpc.server configuration and restart the cluster
[https://issues.apache.org/jira/browse/STORM-2773](https://issues.apache.org/jira/browse/STORM-2773)
There is a cluster which includes three nodes named
storm1,storm2,storm3.And there is a drpcserver in every node,a worker which has
been started on strom1.When strom1 was down with hardware failure,my drpc
topology won't work,when I send request from drpcclient.
As storm1 was down,so the worker will be restarted on another node,but it
can't Initialize successfully because the call method of Adder will throw a
RuntimeException,when drpcspout try to connect to storm1,so the worker will
restart again.
In conclusion,If a drpcserver node in cluster is down,drpc cluster won't
work until we modify the drpc.server configuration and restart the cluster,but
in production,it's difficult to restart whole cluster.
So I think we should catch the RuntimeException and log it,and the drpc
topology will work normally.
You can merge this pull request into a Git repository by running:
$ git pull https://github.com/liu-zhaokun/storm master1010
Alternatively you can review and apply these changes as the patch at:
https://github.com/apache/storm/pull/2365.patch
To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:
This closes #2365
----
commit d3dec0d3ca80db518511bf26810d20035ce0ae81
Author: liuzhaokun <[email protected]>
Date: 2017-10-10T03:05:23Z
[STORM-2773] If a drpcserver node in cluster is down,drpc cluster won't
work if we don't modify the drpc.server configuration and restart the cluster
----
---
