Virtual machine scale set not scaling up on start of cluster

Admin OpenKonsequenz 21 Reputation points
2022-11-08T17:50:48.68+00:00

Hello all,

we use the start stop automation for our clusters. One day it suddenly stop working. By now thanks to the community support and their experts
we can at least start them again
Thread here:
https://learn.microsoft.com/en-us/answers/questions/1052912/subnet-aks-subnet-is-in-use-and-cannot-be-updated.html?childToView=1063878#answer-1063878

The problem is it only works if we manually set the scale of the vmss back to 1 everytime after we start the cluster, else it stays on 0
and of course no instance is starting.

258338-2022-11-08-18-45-34-2022-11-08-18-18-34-aks-nodepo.png

Any ideas.

Best regards
Simon

Azure Virtual Machines
Azure Virtual Machines
An Azure service that is used to provision Windows and Linux virtual machines.
7,123 questions
Azure Virtual Machine Scale Sets
Azure Virtual Machine Scale Sets
Azure compute resources that are used to create and manage groups of heterogeneous load-balanced virtual machines.
347 questions
0 comments No comments
{count} votes

3 answers

Sort by: Most helpful
  1. Manu Philip 16,971 Reputation points MVP
    2022-11-08T19:09:35.877+00:00

    The possible reasons for a scale action did not occur are as follows. Checking these aspects should help in resolving the situation as I believe

    1. There may be no scale action events or logs.
    2. Review the autoscale metrics if you are using a metric-based scale rule. It's possible that the Observed metric value or Observed Capacity are not what you expected them to be and therefore the scale rule did not fire.
    3. It's also possible that the cool-down time kept a scale action from occurring.

    Also, review the autoscale evaluation logs during the time period you expected the scale action to occur. Review all the evaluations it did and why it decided to not trigger a scale action. Following Kutso query can be used:

    AutoscaleEvaluationsLog  
    | where TimeGenerated > ago(2h)  
    | where OperationName == "MetricEvaluation" or OperationName == "ScaleRuleEvaluation"  
    | project OperationName, MetricData, ObservedValue, Threshold, EstimateScaleResult  
    

    ----------

    --please don't forget to upvote and Accept as answer if the reply is helpful--

    0 comments No comments

  2. Admin OpenKonsequenz 21 Reputation points
    2022-11-09T11:28:59.513+00:00

    Thank you @Manu Philip for your quick answer.

    we don't use the autoscale functionality thats why most likely there is no such table like AutoscaleEvaluationsLog.

    1. There may be no scale action events or logs.
      What do you mean by that?
    2. we don't use metric-based scale rule, we only need one vm instance
    3. How to check that and if so how to fix it?

    258734-image.png


  3. Admin OpenKonsequenz 21 Reputation points
    2022-11-09T20:42:03.247+00:00