[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

oo-diagnostics not reporting required dns service status



So I've started up my own instance of openshift origin. I've run into a situation where when creating an application, I'll receive an error message on the web console:

"Unable to complete the requested operation due to: Connection refused - connect(2) (http://192.0.0.100:8053 ). Reference ID:..."

Running oo-diagnostics did not point me in the direction I needed to solve this problem. Looking at the broker production.log, I saw the same error message.

I used the puppet-openshift_origin module to deploy a single instance hosting EVERYTHING. In this case, chosen broker_dns_plugin variable was set to 'avahi', and two services required were not running:

avahi-cname-manager
avahi-daemon

After starting these two services, all was well.

I'm not sure what the recommended dns solution would be, in my particular case I followed the recommendation for my environment which specified the use of avahi. But without those services running, I ended up noticing three things overall:

1. The puppet module doesn't appear to explicitly tell these services to start, [openshift_origin::plugins::dns::avahi](https://github.com/openshift/puppet-openshift_origin/blob/master/manifests/plugins/dns/avahi.pp)
2. Despite the error, the application was still created. This left the app kinda hanging as DNS was never put into place properly.
3. I couldn't delete the application.

I FEEL that if this situation is to occur, the app shouldn't get created. And secondly, oo-diagnostics should get some additions to detect whether or not a DNS update would fail and report to such to the administrator. And thirdly, perhaps the manifest instructing these services to start.

John Skarbek, System Administrator



[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]