🎨🐛Autoscaling: Allow EC2 launches in multiple AvailabilityZones ⚠️ (DevOPS) 🚨 #8210
+1,139
−299
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
sanderegg
added
a:autoscaling
Codecov Report❌ Patch coverage is Additional details and impacted files@@ Coverage Diff @@
## master #8210 +/- ##
==========================================
- Coverage 87.95% 87.92% -0.03%
==========================================
Files 1930 1926 -4
Lines 74773 74762 -11
Branches 1306 1309 +3
==========================================
- Hits 65767 65738 -29
- Misses 8615 8632 +17
- Partials 391 392 +1
Continue to review full report in Codecov by Sentry.
🚀 New features to boost your workflow:
|
sanderegg
force-pushed
the
ec2/allow-multiple-az
branch
5 times, most recently
from
40c92dc to
ae240fa
Compare
sanderegg
changed the title
sanderegg
force-pushed
the
ec2/allow-multiple-az
branch
2 times, most recently
from
03b003f to
9c37d09
Compare
This comment was marked as outdated.
This comment was marked as outdated.
Sorry, something went wrong.
sanderegg
changed the title
sanderegg
force-pushed
the
ec2/allow-multiple-az
branch
from
547d58b to
3be802d
Compare
sanderegg
changed the title
There was a problem hiding this comment.
Pull Request Overview
This PR enables autoscaling services to launch EC2 instances across multiple Availability Zones to improve resilience against capacity limitations. The main purpose is to handle InsufficientInstanceCapacity errors by iterating through multiple subnets in different AZs before failing.
Key Changes:
- Environment variable change from
EC2_INSTANCES_SUBNET_ID(single) toEC2_INSTANCES_SUBNET_IDS(array) for autoscaling and clusters-keeper services - Enhanced EC2 launch logic in aws-library to cycle through multiple subnets when encountering capacity issues
- Addition of new error handling for
EC2InsufficientCapacityErrorandEC2SubnetsNotEnoughIPsError
Reviewed Changes
Copilot reviewed 26 out of 26 changed files in this pull request and generated 6 comments.
Show a summary per file
| File | Description |
|---|---|
| services/docker-compose.yml | Updated environment variable names to use plural form for subnet IDs |
| services/autoscaling/src/simcore_service_autoscaling/core/settings.py | Changed field type from single string to list of strings for subnet configuration |
| services/clusters-keeper/src/simcore_service_clusters_keeper/core/settings.py | Changed field type from single string to list of strings for subnet configuration |
| packages/aws-library/src/aws_library/ec2/_client.py | Enhanced launch_instances to iterate through multiple subnets and handle capacity errors |
| packages/aws-library/src/aws_library/ec2/_errors.py | Added new exception types for insufficient capacity and IP address errors |
| packages/aws-library/tests/test_ec2_client.py | Added comprehensive tests for multi-subnet capacity handling |
| services/autoscaling/tests/unit/test_modules_cluster_scaling_dynamic.py | Added tests for multi-subnet scenarios and warm buffer failure handling |
sanderegg
force-pushed
the
ec2/allow-multiple-az
branch
from
3be802d to
cb6f177
Compare
sanderegg
requested review from
mrnicegyu11,
YuryHrytsuk and
pcrespov
as code owners
Co-authored-by: Copilot <[email protected]>
sanderegg
force-pushed
the
ec2/allow-multiple-az
branch
from
f52dc5a to
db5f577
Compare
|
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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 do these changes do?
This PR:
autoscalingandclusters-keeperservices)InsufficientInstanceCapacity(changes inaws-library), test-driven bytest_ec2_client.pyNOTE1: warm-buffers that were launched to pre-pull images and then stopped at some time, might still fail when needed as the warm-buffer is then hard-linked to the specific AZ where it was launched first. Therefore there is no guarantee that it can be started again at any time.
NOTE2: In this case,
autoscalingshould have coped and tried to launch a fresh instance probably in a different AZ. The additional tests intest_modules_cluster_scaling_dynami.pyshow that this fails and is a pretty big bug, which will need to be fixed: #8273Partial bugfix: if starting warm buffers fail then the autoscaling service will still take care of scaling down and up for other instance types than warm buffer types
Related issue/s
How to test
Dev-ops
EC2_INSTANCES_SUBNET_IDwithEC2_INSTANCES_SUBNET_IDStakes a JSON list of subnet IDs in autoscaling and clusters-keeper (e.g.EC2_INSTANCES_SUBNET_IDS,PRIMARY_EC2_INSTANCES_SUBNET_IDS,WORKERS_EC2_INSTANCES_SUBNET_IDS)