5.1.6

- Add dogstatsd support to statsd monitor

5.1.5

- Added new monitor `ntp` to monitor the clock accuracy of the current host
- Suppression of extraneous error log message that could leak sensitive config
- Fix for span environment tracking stale cache upon agent config restart
- `vsphere` monitor how supports pagination of queries
- Fixed bug where discovery rules with the `python-monitor` monitor would not pass down additional config to the subprocess

5.1.4

- Improve inventory traversal for the vSphere monitor

5.1.3

- Fix expvar monitor to use the metric name option when it's defined

5.1.2

- Upgrade golang to 1.14.2 in the agent build to address https://github.com/golang/go/wiki/LinuxKernelSignalVectorBug.
- Fix a bug in the agent dynamic configuration logic where the `_` config option randomly failed to replace an existing config block.
- Fix a bug in the `telegraf/tail` monitor where the json query failed to parse all entries in the JSON array.
- Introduce a new `cloudMetadataTimeout` flag in the agent config schema to control the HTTP Client timeout for Azure, AWS and GCP. Defaults to 2 seconds.
- Log errors while running AWSUniqueID. Log is only enabled in debug mode.

5.1.1

- Update mongodb and openstack collectd releases

Add service owner option

5.1.0

- Address goroutines leak in docker monitor
- Update initd service script to set ownership for $user variable
- Add skip SSL validation flag for openstack monitor

Breaking Changes:

- The kubernetes-apiserver monitor will no longer emit the apiserver_request_duration_seconds_bucket metric by default

Update min agent version to 5.1.0 for service override

Add service owner option