vROPS 8.16 - Maintenance schedule not working

[srodenburg]

Hello folks,

Bit of a wierd situation here. I need to put one specific ESXi host (out of 90) in vROPS maint.mode (stop monitoring) during 4 hours every sunday.

So I created a Maintenance schedule and a policy which only contains that one host directly. It does not work. The schedule is triggered (checked the start time etc.) but looking at the inventory at that host object, nothing happens. Data collection remain active.

I tried it with a custom group: made one that automatically includes that one host (checked with preview): same thing.

The policy has nothing enabled. So no metrics or anything. Absolutely nothing except the Maintenance schedule. All 90 hosts have a normal monitoring policy assigned and this works fine. This policy is the default policy.

What I am doing wrong. The policy is there, it is active, has Prio 1 and the verified active on that one host. All looks good. Still it does not work and collection is never paused during the Maintenance schedule.

Can it be that having the main (default) policy enabled (with gazillions of Metric etc. etc. enabled) while at the same time having a second policy enabled (with only the Maintenance schedule in it), that they cancel each other out or something?

Cheers,
Steve

[Uday_Mantri]

Hello Steve,

The new maintenance policy that you have created is set on priority 1, has the maintenance schedule set and is applied on the intended host, then it should automatically over-rule the default policy, even if this default policy has that same host.

I would request you to please re-verify if the host has the correct policy assigned. Open the host object and see what is the name of policy at the top right hand corner. There you will also find the option to assign it a policy if you find any deviation, just make sure that your maintenance policy has the top most priority.

Regards,

Uday Mantri

[srodenburg]

Hello Uday,

When I go into the Object Browser -> the vCenter -> the Datacenter object -> Cluster and then select the host and I click "view policy" (or hover over it) it tells me that the correct policy (the maint.policy) is the active policy.

That new Maint.Policy, with Prio "1", has no metrics or anything. It is empty. It only has the Maint.schedule in it. It is now the active policy of this one specific esxi server, instead of the default policy(it has a "D" behind it) that it used to have, and that all other ESXi servers still use (active) and which has all the metrics, super metrics etc. etc. in it.
Question: Did the metrics of the two policies merge, in the sense that this one specific server, which does not have the normal default policy as it's active policy anymore, still have all the metrics from the "Default" policy, together with the content from the Maint.policy? Or does this specific server now have 0 metrics because the Maint.policy has no metrics in it? I hope my question makes sense..

[Uday_Mantri]

I understood the point, basically, the Maintenance Policy which has the schedule, should inherit (should be child of) the other Default Policy. In this way the Maintenance Policy will continue to govern (collect metric and property data) your 1 ESXi Host normally and when the Maintenance schedule kicks in, it would them put that Host in maintenance, meaning stop its data collection. Now, the other important point is that Maintenance Policy should have Priority 1 and the other Polciy should be default.