Alert will not re-trigger after 1 alert per morning.
ABelk
Posts: 4 New member
Hi all,
The main issue I am having is I have an alert set up to run each hour but it will only actually alert once per day. I think the root of the issue is that the alert is being created and places in “Active” status instead of being created in “Ended” status. I feel like an alert will not retrigger if the alert is still active on a specific server being monitored. To provide a little background information on the alert, the alert is looking for any value over zero from a T-SQL standpoint. I think we have that set properly in the threshold section below. Additionally, we have the Frequency set to 1 hour. In this case I would expect 24 alerts per day(1 each hour) if the issue is not resolved. At the moment we seem to be getting 1 alert for each database server around Midnight each day. This makes me feel I have some sort of setting to only allow 1 alert per day or per server until the alert is “ended” for the specific database. I’m sure there is some small setting I might be missing.
Does anyone have an idea of why this might be happening?
Recap:
Frequency: 1 hour
Threshold : 0
Alerts: Created the first time each day.
The main issue I am having is I have an alert set up to run each hour but it will only actually alert once per day. I think the root of the issue is that the alert is being created and places in “Active” status instead of being created in “Ended” status. I feel like an alert will not retrigger if the alert is still active on a specific server being monitored. To provide a little background information on the alert, the alert is looking for any value over zero from a T-SQL standpoint. I think we have that set properly in the threshold section below. Additionally, we have the Frequency set to 1 hour. In this case I would expect 24 alerts per day(1 each hour) if the issue is not resolved. At the moment we seem to be getting 1 alert for each database server around Midnight each day. This makes me feel I have some sort of setting to only allow 1 alert per day or per server until the alert is “ended” for the specific database. I’m sure there is some small setting I might be missing.
Does anyone have an idea of why this might be happening?
Recap:
Frequency: 1 hour
Threshold : 0
Alerts: Created the first time each day.
Tagged:
Best Answer
-
Alex B
Posts: 1,158 Diamond 4
Hi @ABelk
Unfortunately SQL Monitor custom metrics don't currently work in the manner you want and this is expected behavior. The issue here is that the alerts for custom metrics are Continuous type alerts rather than Event type alerts - the Continuous alerts are Active until the condition that triggers the lowest threshold has not been crossed for more than the stated number of collections (e.g. CPU utilization default alert); the event type alerts are "something happened" (e.g. SQL Server Error Log entry alerts from the default SQL Monitor alerts).
You should vote on this SQL Monitor Uservoice suggest.
In the meantime, you will need to design the custom metric to fall below the threshold so it will end so that it can trigger a new alert. Depending on how these numbers greater than zero are behaving (whether increasing, decreasing or static) you may be able to use the rate of change tickbox to get it working as well, but ultimately it's not designed for this specific use-case currently.
I've brought this up with the team now as well but ultimately it's a feature request and so depends on demand vs how much work it is to implement.
I hope that helps clarify the situation!
Kind regards,
Alex
Answers
Thank you for responding. I found this out through testing other alerts within the system. I'll keep this in mind for future alerts!
Thank you for responding. I found this out through testing other alerts within the system. I'll keep this in mind for future alerts!