Metric hacky friday #455
Metric hacky friday #455
Conversation
gertzakis
requested review from
itdependsnetworks,
jeffkala and
nkallergis
as code owners
nautobot_golden_config/metrics.py
Outdated
| backup_gauges = GaugeMetricFamily( | ||
| "nautobot_gc_backup_total", "Nautobot Golden Config Backups", labels=["seconds", "status"] | ||
| ) | ||
| time_delta_to_include = PLUGIN_SETTINGS.get("metrics", {}).get("time_delta", timedelta(days=1)) |
There was a problem hiding this comment.
Is the last day the most common metric? I would have thought it would have been just the status in general
There was a problem hiding this comment.
So what you propose is
success_count = GoldenConfig.objects.filter(backup_last_attempt_date=F('backup_last_success_date')).count()
_attempt_count = GoldenConfig.objects.filter(backup_last_attempt_date__isnull=False).count()
failure_count = _attempt_count - success_count
?
There was a problem hiding this comment.
The status of the last Job run is exposed from the capacity_metrics (which will be migrated into Nautobot core), example visualization here. I like exposing the counts of successful/attempted Job runs within a given interval, because those metrics will be stored in Prometheus and then you may query/process them as you like (e.g. visualize 1 year).
There was a problem hiding this comment.
We had a discussion and agreed that on the operational side of things the metric that Ken proposed would be easier to handle. The time dimension then comes from the time-series database. We just have a gauge that displays the current amount of successful/failing backups/intended/compliance/etc., does that make sense? The key point here is GoldenConfig object status vs. JobResult object status.
There was a problem hiding this comment.
thanks! and much better said than I could have.
There was a problem hiding this comment.
Thank you! I've updated the metrics to reflect your comments. Please check again!
|
Can you address the merge conflicts? We may get one more release out before adding this, still thinking about it. |
|
We are good in principle here, however, just determining which release. |
|
Any plans to release this? |
* POC for Prometheus instrumentation * Advance POC. * add compliance job metric * forgot yield * add simple metric rules per feature * added docstring * clean up stuff * cleanup * update invoke * fix pylint * fix black * fix CI for 1.5.3 for metrics * fix pylint * fix metrics based on comments * fix metrics * update lock file * fix linters * update `poetry.lock` --------- Co-authored-by: Leo Kirchner <leo.kirchner98@gmail.com>
Some basic metrics for Golden Config plugin #453. @Kircheneer helped me and had the baseline for me :)
metric_gc_jobsexposes The successful vs failed GC related jobsmetric_golden_configexposes number of devices that are configured per GC featuremetric_complianceexposes number of compliant vs non-compliant devices per featureBelow there is the exposed metrics in
/metrics/Also, I update a lot of files for passing black, and the
CIto use Nautobot version1.5.13in order to have the metrics functionality.