brharrington

Thanks for the comment. How does Grafana use that setting? > Being able to specify the step size is useful when the metrics being stored are at different intervals than...

If you follow that pattern, then there isn't any reason to use DynamicProperty. Just get the value from the config directly. I think you will have a memory leak with...

For reference, here are some previous discussions on this for other projects: https://github.com/ReactiveX/RxNetty/issues/292, https://github.com/Netflix/ribbon/issues/206

There are a several ways it can be configured. For the [sample config](https://github.com/Netflix/atlas/blob/master/conf/memory.conf) it runs a single node in-memory configuration. I need to finish merging a couple of the internal...

That histogram is odd, there are entries in the middle that are larger than the total at the end: ``` $ cat ~/Downloads/atlas_hist.txt | awk '$3 > 1e9 { print...

> We hit an issue with memory leaks after introducing nginx balancer with mirror module Did anything change other than that? Was it working fine with the same data prior...

> If so, then maybe, Atlas could have an option to fail fast with a simple response. Maybe, it isn't clear that is the cause of the issue though. If...

The total for the new histogram is negative. Looks like that is a known bug that they do not intend to fix: https://bugs.openjdk.java.net/browse/JDK-7012905 https://bugs.openjdk.java.net/browse/JDK-6539434 Did you run `-histo:live` or just...

I have some tests setup and running, should have some results in a few days.

> any non-default value for akka's max-connections causes memory leak after short uptime. Interesting. Did you change that setting when you started using nginx? > Can you confirm that 100...