Andrew Tolbert

Looking into this further, even having clear doesn't seem right. Clearing a meter seems somewhat difficult too..will think on this more.

👍 sounds like a good idea, maybe a top level option in `ScheduledReporter` to control this would be nice.

I think it is not necessary to call it yourself because the [created EWMAs](https://github.com/mikejihbe/metrics/blob/v0.1.11/metrics/meter.js#L6-L8) are [configured](https://github.com/mikejihbe/metrics/blob/v0.1.11/stats/exponentially_weighted_moving_average.js#L57-L59) to tick every 5000ms [by themselves using setInterval](https://github.com/mikejihbe/metrics/blob/v0.1.11/stats/exponentially_weighted_moving_average.js#L19-L20). I've seen in dropwizard metrics that...

This may be possible, although I think the utility of it may be limited. The m1/m5/m15 metrics record rate per second over those time period. Rate over the last second...

I think this is a good idea that I hadn't considered since I primarily use graphite to store my metrics which doesn't have this capability as far as I know,...

Can reproduce it with executable jar too by attaching a yourkit agent. Must just be that there were no full GCs happening in my session.

The result of this is a file stays open per query/page until GC which is bad (such fix it).

This looks very similar to Stratio/cassandra-lucene-index#345 which could have to do with incompatibilities between the version of JNA in use and the version of glibc involved. Will need to investigate...

CASSANDRA-11483 was merged

#59 should solve this, it seems to be working for me from @TimothyGu's branch.