Pool pool state is DEGRADED - alert not repeated

[A210712]

Hello,

On Alert Settings page, at "Pool Status Is Not Healthy" and "Boot Pool Is Not Healthy" I've set "Set frequency" to HOURLY,
but there is only one alert email sent.

I have a mirrored boot pool (2 SSD disks) and a mirrored data pool (2 WD HDD disks). First I detached one boot
disk (in power off state), then I received one alert email. A waited for a couple of hours, nothing happened.

Then I detached one data disk (also in power off state). After booting the system, I received another alert email, at
13:00 (not at boot time), stating as new alert the missing WD HDD disk, mentioning at "Current alerts" the degraded
boot pool.

And that was it. No more alert emails. Am I missing the right setting?

I don't know if it means something, that the Pool state is Degraded part contains the disk model ans serial number
"is UNAVAIL", while at the "Boot pool status is Degraded" part there is no disk mentioned.

Code:

New alerts:
* Pool pool state is DEGRADED: One or more devices could not be opened. Sufficient replicas exist for the pool to continue functioning in a degraded state.
The following devices are not healthy:

    Disk WDC WD5000BEVT-26A0RT0 WD-WXG1A5081101 is UNAVAIL



Current alerts:
* Pool pool state is DEGRADED: One or more devices could not be opened. Sufficient replicas exist for the pool to continue functioning in a degraded state.
The following devices are not healthy:

    Disk WDC WD5000BEVT-26A0RT0 WD-WXG1A5081101 is UNAVAIL


* Boot pool status is DEGRADED: One or more devices could not be used because the label is missing or invalid. Sufficient replicas exist for the pool to continue functioning in a degraded state..

 

[sretalla]

And that was it. No more alert emails. Am I missing the right setting?

You didn't see an email saying the alert was cleared did you?... So I guess you want it to SPAM you into action to fix it then? That's not how it works.

The frequency is how often it checks if there's a new alert to send, not how often to send the same alert over and over again.

Booting the system assumes no messages are sent yet, so messages are potentially "repeated" in that case.

Those are 2 different errors you have, so the different formatting is a reflection of that.

 

[A210712]

You didn't see an email saying the alert was cleared did you?... So I guess you want it to SPAM you into action to fix it then? That's not how it works.

The frequency is how often it checks if there's a new alert to send, not how often to send the same alert over and over again.

Booting the system assumes no messages are sent yet, so messages are potentially "repeated" in that case.

Those are 2 different errors you have, so the different formatting is a reflection of that.

The documentation says about "Set frequency": "... how often alert notifications are sent ."

And that's exactly what I want, that is to receive hourly an e-mail if there is an ongoing critical condition.

 

[sretalla]

The documentation says about "Set frequency": "... how often alert notifications are sent ."

Although it's apparently not clear enough in saying that only new or cleared alerts are sent at those intervals.

I invite you to observe reality (that you have already described on your system... and is the same on mine) as the evidence for that.

 

[A210712]

That is called "hourly digest". That might work too.
What I'm still looking for is a setting, to send that hourly digest regardless of there are new alerts or not. Do you know any setting for that?

 

[sretalla]

Do you know any setting for that?

I don't think there is one. You may need to submit a feature request in Jira and get some people to vote for it.