AVD Scaling Logic Duplicate Trigger

Avatar
Tomas

Currently you can set autoscale to trigger based on CPU, Ram, or based on session where you can say for example if CPU gets to 60% for 15 minutes add a host and then add another trigger with a different resource as the criteria

We have a client that has some weird CPU usage and it brought up having 2 triggers for CPU.
For example: If CPU exceeds 70% for 30 minutes OR if CPU exceeds 90% for 5 minutes.

This would let us have a bit of flexibility for these clients who can spike their CPU for several minutes but then would otherwise have little usage.

2

Comments (3 comments)

Sort by
1
Avatar
Dave Stephenson

I can't say I've seen this asked for before, but it sounds intriguing.

If I understand correctly, you want something (like the screenshot below) where you can have multiple CPU (or RAM) triggers based on your thresholds.
I can see that feature being difficult to troubleshoot for our support team if an issue does arise. But if our Dev team can pull something off like this, it could add quite a bit of flexibility.

Out of curiosity, for your use case, would you need more than 2 triggers of each kind, or would that be sufficient?
Is there a scenario where it would need to get super complicated (IF .... CPU %1 = ## AND CPU %2 = ## THEN ... ELSE . . . .) or could we still with the ANY condition for the Scale Out?

0
Avatar
Tomas

No what you are showing there is pretty much what we would want. just a simple or statement. 

0
Avatar
Dave Stephenson

Excellent!
Simpler does make the likelihood of getting it implemented easier. 🙂

Please sign in to leave a comment.