Merge pull request #50 from ashangit/master

 maxAgeSeconds has badly been increase which leads to false positive latency issue

Increase number of bucket and age of Latency memcached summary

Currently we only kept one bucket which can lead to NaN value just after the roll of bucket.
Keeping more bucket to avoid this situation as we also have scrape value at 1 min it ccould leads to
almost always NaN value beeing scrapped by prometheus

Run ConsulDiscovery in a separate thread from poke one

Since we contact each couchbase cluster to get real list of nodes
the ConsulDiscovery takes too much times which affect number of time we retrieve metrics

On consul discovery query couchbase node only for couchbase service t…

…ype (#47)

Add gradlew

Create releases.yml

Update consul discovery mechanism to get list of nodes from couchbase (…

…#45)

List of nodes can be updated on consul side but not on the couchbase cluster.
Due to that if we recreate the monitor from the list of nodes in consul we can still have a monitor querying a node that will be removed.
Then when it is really rermoved from couchbase we are not recreating the monitor which leads to some prometheus metrics from this removed node to
be still here.
With that change we will get list of nodes from consul and then query couchbase from thise nde list to get the list of nodes frm couchbase point of view

Create build.yml

Remove disk latency due to too many false positive

Update grgit version and use tag instead of commit for version