Has anybody solved the docker hostname problem?

[Timothy Ehlers]

We are having a big problem with how docker + dropwizard + hostname are setup.

Most users either use the hostname or maybe dropwizard does it by default. The problem comes when you use Mesos/ECS whatever and a container has a problem and gets restarted repeatidly.

For example in 7 days one of my apps restarted 110,000 times filling my graphite disk.

stats.gauges.foo-app.asdb78213as.metric.foo

stats.gauges.foo-app.haeer78213as.metric.foo

stats.gauges.foo-app.zdfadfs3as.metric.foo

stats.gauges.foo-app.earer78213as.metric.foo

stats.gauges.foo-app.aasdf78213as.metric.foo

stats.gauges.foo-app.bhnm8213as.metric.foo

What i'd like to do is something like this, maybe using service discovery to see what other hosts are up and just naming yourself server1/2/3/4/5 based on the count:

stats.gauges.foo-app.server1.metric.foo

stats.gauges.foo-app.server2.metric.foo

stats.gauges.foo-app.server3.metric.foo

stats.gauges.foo-app.server4.metric.foo

stats.gauges.foo-app.server5.metric.foo

Has anybody solved this problem? and yes i know the first problem is the app shouldn't restart 110k times in 7 days.

[Erik van Oosten]

We pass the hostname to use for metrics via an environment property.

Kind regards,
    Erik.

Op 09-11-16 om 16:48 schreef Timothy Ehlers:

--
You received this message because you are subscribed to the Google Groups "metrics-user" group.
To unsubscribe from this group and stop receiving emails from it, send an email to metrics-user...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.