Set the minInstancesInServiceParameters
property
#548
Closed
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
What does this change?
This is to allow testing of guardian/riff-raff#1383 whilst the
riff-raff.yaml
generator from GuCDK hasn't been updated.This
riff-raff.yaml
is a stripped down version of the generated one, only uploading the application artifact to S3, and deploying the application CloudFormation stack.How to test?
This branch can test multiple scenarios.
Note
All scenarios match previously observed behaviour regarding number of in service instances.
"Normal" capacity (i.e.
desired = min
)With the ASG operating at normal capacity, the deployment was also normal.
The deployment log also shows the value of the
MinInstancesInServiceForcdkplayground
CFN parameter was, expectedly, set to match the desired capacity (1).Partially scaled (i.e.
min < desired < max
)Before a deployment was started, the
scale-out
script was executed a few times, causing the ASG to havemin=1
,desired=5
,max=10
.The deployment was normal and the deployment log shows the value of the
MinInstancesInServiceForcdkplayground
CFN parameter was, expectedly, set to match the current desired capacity (5):The metrics show the ALB's healthy host count never dropped below 5. That is, there was no degradation in service availability.
Fully scaled (i.e.
desired = max
)Before a deployment was started, the
scale-out
script was executed a few times, causing the ASG to havemin=1
,desired=10
,max=10
.The deployment was predictably slow as it operated as a "one out, one in" update. The deployment log shows the value of the
MinInstancesInServiceForcdkplayground
CFN parameter was, expectedly, set to 7 (75% of max, rounded down):The metrics show the ALB's healthy host count dropped to 7, as expected. Whilst this results in a drop in availability, it is an unusual scenario as scaling to max is indicative of incorrect capacity planning. That is, the max should be higher.