Patchmon Host Monitoring Alerts
Since Patchmon has already all host system data, it would be beneficial to extend the alerts and add monitoring notifications for disk, cpu and memory usage. It should be possible to maintain the thresholds for each in the application.
6 Comments
I strongly recommend against it. This would put PatchMon on the slippery slope of turning into a general purpose infrastructure monitoring system. It might be tempting to work on data easily available. But for creating a proper "disk full soon" check, you have to calculate rates over the last hours to determine an "hours until full" metric (which says much more than "disk is 99% full", that often is no problem on a 50TB array, while fast accumulating logs on a 240GB SSD at 70% full can be a massive problem).
So, I'd really suggest to use a 'classical' infrastructure monitoring tool like Checkmk, Icinga, Nagios or Zabbix (alphabetical order) to do this.
And: yes, I am employee of Checkmk. But many of our users have identified PatchMon as their ideal complementary monitoring tool since PatchMon does patch monitoring right. These users really like PatchMon since it is a unified interface for all covered operating systems to the different approaches of patch monitoring that are quite fragmented over the OS families in the monitoring platforms mentioned before. IMHO development of PatchMon should concentrate on patch monitoring and keep overlap to infrastructure monitoring tools low. And I will do my best for building PatchMon integrations for classical infrastructure monitoring platforms.
Hi Mattias, I fully understand your concerns. It was just an idea to add a simple monitoring, since all data is already available but I agree to your concern. This can be closed.
I'd be inclined to agree - but as a good compromise we could do the following:
- If integrations with Zabbix or any of the other monitoring solutions exist, then link to those from the host view in PatchMon.
I think that would be the best of both worlds.
See this feature request. Currently, I can create links to PatchMon from the service list, but not yet from the host list.
Nice - upvoted!
Hi team. It can be very tempting to expand our current feature set into monitoring other aspects of the OS but it's a minefield of a task.
Our aim in PatchMon is to build a smart Patching solution to cover security and whilst a lightweight metric on resources is available - it's not the primary aim.
Internally we use Zabbix and now working with Checkmk and because they can monitor things like switches and other types of data (Http, snmp etc) then we feel like it will cause a drift and duplication.
The exact case is where someone wants to use Zabbix for managing Patches in Linux (it works the other way round) but by having integrations into Zabbix, Checkmk (and also i'm soon to be working on Grafana) then it allows for individual tools to be used in full capacity without overlap.
Hope that helps - keep up the bringing of feature requests though, as they always allow us to think and re-align on our promise.