Instance Discovery Grace Period Issue

[Tursiops]

Hi,

I noticed that the new feature of a grace period for DCIs sets instances to disabled when they are no longer found.
However, it does not appear to re-activate them when they re-appear, which means unless one manually goes through the nodes to locate such DCIs to re-activate them, they will simply be deleted after the grace period.

Not sure if that is the intended behaviour?

Cheers

[Sviests]

Hi!

Could you run instance discovery on the node manually and post a screenshot of the output?

[Tursiops]

Hi,

Of course since I posted I have not run into this again.
The only ones that are currently staying disabled are indeed due to OID changes or similar events which trigger new instances.
I'll let you know if/when I encounter this again.

Cheers

[Tursiops]

Hi,

Just found one and attached a screenshot, which basically shows that nothing happened while updating instances.
Yet, the "Hyper-V" ones had a number of Disabled instances.
For all I can tell Instance Discovery ran at some point while the relevant service on the server wasn't up, so it couldn't generate the instance list at the time. This resulted in some garbage instances (I need to add better error handling to the external list command). Either way, those garbage instances were properly disabled once the service was back up. However, the correct instances were not re-enabled. The moment I manually switched them to "Active", they just started working without any further changes necessary.

Cheers

---

[Sviests]

Hi,

If you have debug level 5 enabled, please check your server logs for a message that is similar to this: "DataCollectionTarget::updateInstances(%s [%u], %s [%u]): instance \"%s\" not found, grace period started" without the "%s and %u symbols. This should indicate that the grace period was actually started as it is what triggers the re-enabling of the instance. It could be an issue that the instance went into an unsupported state for a different reason and was not re-enabled by instance discovery.

[Tursiops]

I generally keep debug to level 1 (or off) unless something is wrong, as I otherwise end up with several gigabytes of logs in a matter of minutes.
Is there a way to set debug level to 5 for this particular area only?

[Sviests]

We are starting to implement tags for debug messages so that particular debug messages could be filtered, but unfortunately this has not been done yet for the messages in question.