frankjkelly opened a new pull request, #391:
URL: https://github.com/apache/pulsar-helm-chart/pull/391

   …scaleDown.
   
   ### Motivation
   
   While testing Pulsar Broker autoscaling we found that during a scale-down 
event (going from 14 brokers to 2 in the space of about 15 minutes) that 
namespace bundle movement occurred very frequently and we would like to slow 
the rate at which the brokers scale down using Scaling Policies.
   
   See 
https://kubernetes.io/docs/tasks/run-application/horizontal-pod-autoscale/#configurable-scaling-behavior
 for more details
   
   We would also like to use the `stabilizationWindowSeconds` to ensure that 
if/when we add brokers we don't suddenly scale-down soon after (sometimes CPU 
spikes can occur as brokers die, new brokers launch and that can result in a 
temporary spike in average broker CPU causing scale-up. When the spike fades 
then it may cause a scale-down resulting in namespace bundles moving yet again).
   
   ### Modifications
   
   Provide hooks in the `hpa` helm chart templates to provide behaviors e.g. 
   
   ```
   behavior:
     scaleDown:
       stabilizationWindowSeconds: 600
       policies:
       - type: Pods
         value: 1
         periodSeconds: 60
   ```
   
   ### Verifying this change
   
   - [ ] Make sure that the change passes the CI checks.
   


-- 
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.

To unsubscribe, e-mail: [email protected]

For queries about this service, please contact Infrastructure at:
[email protected]

Reply via email to