Adding optional service_check_rw parameter to check for read-only filesystem #2086
Conversation
There was a problem hiding this comment.
Hey @bberezov, thanks a lot for your contribution. Looks like a nice feature to add to the check.
I just have a few comments. Would you mind addressing them ?
Also, could you please document the new service check in a service_check.json file (see docs for details), as well as in README.md under the Service Checks section (see here for an example) ?
disk/CHANGELOG.md
Outdated
| @@ -1,5 +1,9 @@ | |||
| # CHANGELOG - disk | |||
|
|
|||
| ## 1.3.0 / 2018-08-21 | |||
There was a problem hiding this comment.
No need to update the changelog manually, this will now automatically be taken care of at release time
| # | ||
| # The (optional) service_check_rw parameter notifies when one of the partitions is in | ||
| # read-only mode | ||
| # service_check_rw: no |
There was a problem hiding this comment.
Can you put false for the default value here as per #2078 ?
disk/datadog_checks/disk/disk.py
Outdated
| @@ -127,6 +129,16 @@ def collect_metrics_psutil(self): | |||
| self.gauge(metric_name, metric_value, | |||
| tags=tags, device_name=device_name) | |||
|
|
|||
| # Add in a disk read write or read only check | |||
| if self.instances[0].get('service_check_rw', False): | |||
There was a problem hiding this comment.
you should use self._service_check_rw that you set here https://github.com/DataDog/integrations-core/pull/2086/files#diff-4d0d0ed796595d875c2f4710bb59b678R63
disk/tests/test_check.py
Outdated
| c = Disk('disk', None, {}, instances) | ||
| c.check(instances[0]) | ||
|
|
||
| aggregator.assert_service_check('disk.read_write', status=2) |
There was a problem hiding this comment.
You can use the constant Disk.CRITICAL for the status.
disk/datadog_checks/disk/disk.py
Outdated
| # Add in a disk read write or read only check | ||
| if self.instances[0].get('service_check_rw', False): | ||
| rwro = list(set(['rw', 'ro']) & set(part.opts.split(','))) | ||
| if len(rwro) == 1: |
There was a problem hiding this comment.
Should we send a service check with AgentCheck.UNKNOWN status when neither rw nor ro is there ?
disk/datadog_checks/disk/disk.py
Outdated
| self.service_check( | ||
| 'disk.read_write', | ||
| AgentCheck.OK if rwro[0] == 'rw' else AgentCheck.CRITICAL, | ||
| tags=tags+['device_name:%s' % (device_name)] |
There was a problem hiding this comment.
The tag key here should be device.
|
Hi @zippolyte I've updated PR according to your comments, let me know if anything else has to be modified. |
disk/service_checks.json
Outdated
| @@ -0,0 +1,11 @@ | |||
| [ | |||
| { | |||
| "agent_version": "5.22.3", | |||
There was a problem hiding this comment.
This should be 6.5.0 since it will be released in the next version of the agent.
There was a problem hiding this comment.
sorry didn't know what that version was for
What does this PR do?
This PR adds read-write/read-only service check for disk.py which notifies if any of the disks are in read-only state.
Motivation
We need this functionality in our production environment in addition to standard disk metrics in order to catch any of the disks falling back to read-only mode.
Review checklist
no-changeloglabel attachedAdditional Notes
The test for this change passes our local testing environment replicated from TravisCI configuration