[
https://issues.apache.org/jira/browse/AURORA-840?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14173184#comment-14173184
]
Dobromir Montauk commented on AURORA-840:
-----------------------------------------
I'm running mesos-master with:
/usr/local/sbin/mesos-master
--zk=zk://10.120.218.80:2181,10.88.29.88:2181,10.88.26.227:2181/mesos
--port=5050 --log_dir=/var/log/mesos --quorum=1 --work_dir=/var/lib/mesos
So that was the problem. Thanks!
A cleaner error message would be nice. There are 4 scenarios we should
distinguish:
1. Cannot connect to Zookeeper to get the ServerSet at all.
2. Connected to Zookeeper, but the ServerSet is empty (i.e. wrong path, or
Mesos isn't running).
3. ServerSet returns filled, but unable to connect to any of the IPs.
4. Connected to the IPs, but still cannot register.
Having a custom error message for each would make debugging a *lot* easier.
> Add an FAQ for cluster operators
> --------------------------------
>
> Key: AURORA-840
> URL: https://issues.apache.org/jira/browse/AURORA-840
> Project: Aurora
> Issue Type: Story
> Components: Documentation
> Reporter: Bill Farner
>
> There are a number of common stumbling blocks to getting a new cluster
> running, we should start documenting them to help folks self-serve. Off the
> top of my head:
> - Replicated log not initialized
> - Scheduler not registered
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
