Add experimental event ruler #10615

joe4dev · 2024-04-08T17:25:53Z

Motivation

We have several parity gaps and bugs in multiple different implementations for event pattern matching (see PR #10599). We can offer high AWS-parity quickly using the Amazon EventBridge Event Bus event matching now runs on Open Source Event Ruler v1.5.0 released event-ruler Java library from AWS described in this blog post.

Changes

Add an experimental opt-in EVENT_RULE_ENGINE=java flag using the Java event-ruler library from AWS through JPype.
Add a MavenPackageInstaller utility ~~(could probably move to localstack/packages/core.py DONE)~~

Testing

Set EVENT_RULE_ENGINE=java
Ensure that JAVA_HOME points to a valid JVM installation if you are testing in host mode
Uncomment all SKIP_LABELS for the test test_test_event_pattern
Run the test test_test_event_pattern

LocalStack Shutdown

@dfangl and @joe4dev tested the LocalStack shutdown as follows:

Build LS image using make docker-build
Start LocalStack using EVENT_RULE_ENGINE=java DEBUG=1 localstack start
Run a test case for event pattern matching (e.g., `tests.aws.services.events.test_event_patterns.test_test_event_pattern_with_escape_characters) against the running LS instance
docker stop $CONTAINER_ID should stop the container immediately and not after a 10s delay

=> worked without delay

Limitations

The cross-language call from Python to Java is fast but could add a few milliseconds of delay.
The tight coupling between Python and Java sharing the same process could crash each others language runtime upon unexpected errors.
We had to disable the JVM shutdown because LocalStack has too many daemon threads running and JPype only triggers the shutdown after all daemon threads terminated to prevent shutting down the JVM while potentially still in use.
The arm build failed with ERROR: Could not build wheels for jpype1 due to gcc: fatal error: cannot execute ‘cc1plus’: execvp: No such file or directory. Supposedly, JPype only supports up to Python 3.10 🤔, which would be a blocker 🔴 (they recently merged a fix for Python 3.12 though). We are using Python 3.11 in the image and jpype1 builds fine on my ARM MacBook using Python 3.11.7.
- This build error only happened for ARM builds on Linux and required the extra dependency apt-get install -y g++.
- For S3, we had to add this installation step into the make target docker-run-tests-s3-only to keep it out of the S3 image.

Kinesisanalytics

Kinesisanalytics seems incompatible with EVENT_RULE_ENGINE=java indicated by the following failing tests:

TestKinesisAnalyticsProvider.test_tag_list_tag_untag_resource
TestKinesisAnalyticsProvider.test_input_processing_configuration
TestKinesisAnalyticsProvider.test_application_output
TestKinesisAnalytics.test_run_siddhi_query

⇒ all these tests keep failing in the AMD64 pipeline. They are not ARM compatible.
⇒ They fail with the assertion: AssertionError: assert 'STARTING' in ['READY', 'RUNNING']
⇒ The siddhi query test also fails with a name error NameError: name 't_siddhi_api_core' is not defined in the Siddhi loader PySiddhi/SiddhiLoader.py

LS uses some badly outdated fork of the unmaintained (5y; LS 3y last commit) PySiddhi project: https://github.com/localstack/PySiddhi. They are using some kind of custom Python⇒Java bridge based on JNI, which could cause interference.

ℹ️ We decided to ignore this limitation for now. If JPype proves successful, it might be a better alternative than PySiddhi ...

TODO

Switch default EVENT_RULE_ENGINE back from "java" (used for CI testing) to ""
Clarify dev dependencies for host mode docs#1182

Follow Up

Fix LS Python-native implementation for tests.aws.services.lambda_.test_lambda_integration_dynamodbstreams.TestDynamoDBEventSourceMapping.test_dynamodb_event_filter
- numeric_filter fails due to missing anything-but support
- exists_false_filter fails due to broken implementation
- => It is likely better to adopt Ben's SNS implementation and unify them all because the current status is hardly maintainable.

github-actions · 2024-04-09T14:38:24Z

S3 Image Test Results (AMD64 / ARM64)

2 files ±0 2 suites ±0 2m 56s ⏱️ -8s
398 tests ±0 346 ✅ ±0 52 💤 ±0 0 ❌ ±0
796 runs ±0 692 ✅ ±0 104 💤 ±0 0 ❌ ±0

Results for commit dc265aa. ± Comparison against base commit 4b72d6e.

♻️ This comment has been updated with latest results.

github-actions · 2024-04-11T22:24:03Z

LocalStack Community integration with Pro

2 files ± 0 2 suites ±0 1h 30m 58s ⏱️ -12s
2 882 tests +32 2 598 ✅ +32 284 💤 ±0 0 ❌ ±0
2 884 runs +32 2 598 ✅ +32 286 💤 ±0 0 ❌ ±0

Results for commit dc265aa. ± Comparison against base commit 4b72d6e.

This pull request removes 35 and adds 67 tests. Note that renamed tests count towards both.

tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApi ‑ test_authorizer_crud_no_api
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApi ‑ test_create_proxy_resource
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApi ‑ test_create_proxy_resource_validation
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApi ‑ test_create_resource_parent_invalid
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApi ‑ test_create_rest_api_with_custom_id_tag
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApi ‑ test_create_rest_api_with_optional_params
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApi ‑ test_create_rest_api_with_tags
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApi ‑ test_delete_resource
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApi ‑ test_doc_arts_crud_no_api
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApi ‑ test_get_api_case_insensitive
…

tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApiAuthorizer ‑ test_authorizer_crud_no_api
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApiDocumentationPart ‑ test_doc_parts_crud_no_api
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApiMethod ‑ test_method_lifecycle
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApiMethod ‑ test_method_request_parameters
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApiMethod ‑ test_put_method_model
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApiMethod ‑ test_put_method_validation
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApiMethod ‑ test_update_method
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApiMethod ‑ test_update_method_validation
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApiModels ‑ test_model_lifecycle
tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApiModels ‑ test_model_validation
…

This pull request removes 1 skipped test and adds 1 skipped test. Note that renamed tests count towards both.

tests.aws.services.apigateway.test_apigateway_api.TestApiGatewayApi ‑ test_get_api_case_insensitive

tests.aws.services.events.test_event_patterns ‑ test_test_event_pattern[exists_dynamodb_NEG]

♻️ This comment has been updated with latest results.

joe4dev · 2024-04-12T08:21:54Z

localstack/config.py

+# Determine which implementation to use for the event rule / event filtering engine used by multiple services:
+# EventBridge, EventBridge Pipes, Lambda Event Source Mapping, SNS
+# Options: provider (default) | java
+EVENT_RULE_ENGINE = os.environ.get("EVENT_RULE_ENGINE", "").strip()


✅ CI run with JPype enabled using java here in 017bae5

joe4dev · 2024-04-12T08:22:27Z

localstack/services/events/event_ruler.py

+ # Workaround to unblock LocalStack shutdown. By default, JPype wait until all daemon threads are terminated,
+ # which blocks the LocalStack shutdown upon any leaked thread (know LS issue).
+ # See https://github.com/MPh-py/MPh/issues/15#issuecomment-778486669
+ jpype_config.destroy_jvm = False


That's unfortunately needed. Could this cause any side-effects?

The container shutdown will probably not work, as we have an orphan process which did not get terminated by its parent on shutdown. The supervisor will probably wait forever, until docker kills the container.

joe4dev · 2024-04-12T08:25:21Z

pyproject.toml

@@ -67,6 +67,8 @@ base-runtime = [
 "Werkzeug>=3.0.0",
 "xmltodict>=0.13.0",
 "rolo>=0.4",
+ # allow Python programs full access to Java class libraries. Used for opt-in event ruler.
+ "JPype1>=1.5.0",


431KB without any extra dependencies (see their pyproject.toml), just packaging and typing_extensions, which we have as well

could we move it to runtime instead? base-runtime is the base one used by the S3 image, which has no dependency basically. Or was there a specific reason to put into base-runtime?

For a bit of context, basically, base-runtime is the minimal set of dependencies needed to run a service in LocalStack which would be entirely written in Python with no external dependencies (like S3).
I believe this change, which is for a provider with external dependencies and only opt-in, would fit better in runtime.

This would also have the side-effect of not having to change the S3 image Dockerfile, which does not need this and g++.

That makes a lot of sense @bentsku 💯 Thank you for this valuable input 👀 👏
I fixed everything (including the follow-up S3 tests because the test deps include the runtime deps 😬 ; that's a bit of an ugly hack though)

I didn't understand the full implications of base-runtime vs runtime dependencies.

With the fixed solution, the S3 image does not get larger 👌

joe4dev · 2024-04-12T08:30:03Z

tests/aws/services/lambda_/test_lambda_integration_dynamodbstreams.py

+ # {"id": {"S": "id_value_1"}},
+ # {"id": {"S": "id_value_2"}, "presentKey": {"S": "presentValue"}},
+ # {"dynamodb": {"NewImage": {"presentKey": [{"exists": False}]}}},
+ # 2,


This false negative took me forever to spot. A combination of bad test design and inappropriate configuration led to wrong assumptions!
It was configured to 1 but both events match against AWS. However, the second event never gets waited. Negative testing for the last event is unreliable!

Will this test be reintroduced at a later point?

joe4dev · 2024-04-12T08:32:38Z

tests/aws/services/lambda_/test_lambda_integration_dynamodbstreams.py

 function_name = f"lambda_func-{short_uid()}"
 table_name = f"test-table-{short_uid()}"
 max_retries = 50

 create_lambda_function(
 handler_file=TEST_LAMBDA_PYTHON_ECHO,
 func_name=function_name,
- runtime=Runtime.python3_9,
+ runtime=Runtime.python3_12,


/cc @Morijarti I'm bumping the runtime here together with significant test changes, so can can simply take over my changes upon rebasing

Awesome, thank you

joe4dev · 2024-04-12T08:33:43Z

tests/unit/services/lambda_/test_lambda_utils.py

@@ -136,7 +136,7 @@ def test_no_match_partial(self):
 filters = [
 {
 "Filters": [
- {"Pattern": json.dumps({"partitionKey": "2", "data": {"City": ["Seattle"]}})}
+ {"Pattern": json.dumps({"partitionKey": ["2"], "data": {"City": ["Seattle"]}})}


another example why non-validated unit-tests are tricky to get right 😢

joe4dev · 2024-04-12T08:35:11Z

Dockerfile.s3

- apt-get install -y gcc
+ apt-get install -y gcc && \
+ # Workaround to fix JPype1 compile error on ARM Linux "gcc: fatal error: cannot execute ‘cc1plus’"
+ apt-get install -y g++


I wished this was not necessary, especially not needed for S3 😿

I guess this would be a bit faster if it would be in a single apt-get install call? 🤔

And maybe this should be added somewhere in the contribution guide? Because it seems Mac users will struggle to run make install without g++ (and Java when they use this feature) on their machines?

+1 for a single apt-get call, updated ✅

The g++ dependency might be specific to ARM Linux because it worked out of the box for me on an ARM MacBook.

I agree that the Java installation and setting JAVA_HOME is a bit cumbersome in host mode 😿
@alexrashed Where is the right place to document service-specific requirements for host mode? (we have Notion, the dev guide on the website, getting started guide)

I think this should be part of the contribution guidelines (because it affects all open source contributors), which is currently located in our docs.

Why do we need it for S3? Because we put it into base-runtime? Could we just move the dependency into runtime instead? I'm not sure I entirely follow here.
(I probably should add the S3 docker file into the code owners list for S3 as well, I totally missed this PR 😅)

We don't need it for S3. Thanks for suggesting the switch to runtime deps 👍
Fixed as commented in #10615 (comment)

alexrashed

Wow, this a pretty sophisticated change which drastically improve parity with the event pattern matching! 💯
I hope the new dependency doesn't cause too many issues with #10235 (but as you stated, they seem to work on compatibility above 3.10).
Besides that I only had some thoughts concerning the internal developer experience (Mac users need to be aware that they have to install g++) and some smaller nitpicks ;)
Besides that, I'll leave the final signoff to the service owners :)

localstack/services/events/packages.py

alexrashed

Awesome! 🚀 Thanks a lot for addressing all the comments! From my side all these changes are looking good! As mentioned in the initial comment, I'll leave the final sign-off to the code owners 🥳

localstack/services/events/provider.py

dfangl · 2024-04-15T13:19:02Z

localstack/services/events/event_ruler.py

+ # Workaround to unblock LocalStack shutdown. By default, JPype wait until all daemon threads are terminated,
+ # which blocks the LocalStack shutdown upon any leaked thread (know LS issue).
+ # See https://github.com/MPh-py/MPh/issues/15#issuecomment-778486669
+ jpype_config.destroy_jvm = False


The container shutdown will probably not work, as we have an orphan process which did not get terminated by its parent on shutdown. The supervisor will probably wait forever, until docker kills the container.

localstack/services/lambda_/event_source_listeners/utils.py

localstack/services/events/event_ruler.py

dfangl · 2024-04-15T13:23:02Z

localstack/services/events/event_ruler.py

+if config.EVENT_RULE_ENGINE == "java":
+ import jpype
+ import jpype.imports
+ from jpype import config as jpype_config
+ from jpype import java
+


Is it necessary to do this on module level on import? including the startup? Could it be done on first match as well?

I guess we can go full dynamic import upon first event matching:

it could hide potential issues

JVM startup delay is probably not an issue though.

Do you think a lazy import is better here?

Resolved using a dynamic import now (as discussed in sync).

dfangl · 2024-04-15T13:25:02Z

tests/aws/services/lambda_/test_lambda_integration_dynamodbstreams.py

+ # {"id": {"S": "id_value_1"}},
+ # {"id": {"S": "id_value_2"}, "presentKey": {"S": "presentValue"}},
+ # {"dynamodb": {"NewImage": {"presentKey": [{"exists": False}]}}},
+ # 2,


Will this test be reintroduced at a later point?

localstack/services/lambda_/event_source_listeners/dynamodb_event_source_listener.py

joe4dev · 2024-04-15T16:18:57Z

Trying to reply to @dfangl comment: #10615 (comment)

The container shutdown will probably not work, as we have an orphan process which did not get terminated by its parent on shutdown. The supervisor will probably wait forever, until docker kills the container.

This a problem ⚠️ 😢

How should we test this? Any suggestion on how to fix this?

joe4dev · 2024-04-16T14:32:00Z

Trying to reply to @dfangl comment: #10615 (comment)

The container shutdown will probably not work, as we have an orphan process which did not get terminated by its parent on shutdown. The supervisor will probably wait forever, until docker kills the container.

This a problem ⚠️ 😢

How should we test this? Any suggestion on how to fix this?

This should be resolved after successfully testing the shutdown behavior together with @dfangl (in sync today)

The CI is looking good here 🟢 . In ext, we have some Kinesis analytics test still failing (e.g., TestKinesisAnalyticsProvider.test_application_output) ....

joe4dev · 2024-04-17T07:59:52Z

Using EVENT_RULE_ENGINE=java, all tests are 🟢 in community. However, in ext, we have some Kinesis analytics test still failing (e.g., TestKinesisAnalyticsProvider.test_application_output).

The failing Kinesisanalytics tests are likely caused by an interaction with our Kinesisanalytics implementation using https://github.com/localstack/PySiddhi, an unmaintained project that uses some kind of custom Python=>Java bridge. This C-based proxy uses JNI (like JPype) and could potentially cause an unintended interaction.

Failing ext test run:

TestKinesisAnalyticsProvider.test_tag_list_tag_untag_resource
TestKinesisAnalyticsProvider.test_input_processing_configuration
TestKinesisAnalyticsProvider.test_application_output
TestKinesisAnalytics.test_run_siddhi_query

joe4dev · 2024-04-17T10:25:03Z

Without using the new opt-in Java-based event-ruler, all tests are green 🟢 , both in community and ext 🚀

dfangl

Looks good now, thank you for addressing my comments and avoiding start of the jvm on module import.
Looking forward for more unification of all the event matching implementations :)

andy-g · 2024-04-17T14:24:28Z

Could this also be supported when running localstack in a docker container? And I assume this will resolve the current issues with what appears to be unsupported syntax such as:

$or matching
exists matching

There are current matching rules that don't appear to be working for me, I'm currently running localstack 3.3 in docker.

joe4dev · 2024-04-17T15:06:42Z

Thanks for the comment @andy-g

Could this also be supported when running localstack in a docker container?

Yes, it will work in Docker.

And I assume this will resolve the current issues with what appears to be unsupported syntax such as:
Correct, I assume EVENT_RULE_ENGINE=java should fix all issues related to event pattern matching for EventBridge, EventBridge Pipes, and Lambda Event Source Mapping. It should reach almost 100% AWS parity. We currently do not yet implement the same complexity calculation, so very complex pattern might work in LocalStack but throw an exception in AWS.
We are also looking into how to adopt the same rule engine for SNS (which already implements some complexity calculations). //cc @bentsku

All builds are 🟢 , so I expect this to be merged anytime soon 🚀

andy-g · 2024-04-17T15:13:05Z

All builds are 🟢 , so I expect this to be merged anytime soon 🚀

Brilliant, thanks @joe4dev, I'll keep an eye on this PR then, and thanks so much for the quick response.

maxhoheiser · 2024-04-17T18:34:21Z

Very nice PR, I expect to use this heavily in our new native event bridge implementation. Although tests with this java pattern matching take a bit longer to execute for me compared to our own implementation - it is much more complete. I also expect it to perform much better once a certain event number is put through the system at the same time. This will come in handy if we can get the rest of the provider's performance also up that level of speed.

bentsku

I've added 2 comments: #10615 (comment) and #10615 (comment), which I believe should be addressed before merging if possible, as this PR would then have no impact on the S3 image or base runtime. There might be something I missed, feel free to correct me if it's really needed to go into base-runtime. Thanks!

This reverts commit 4e7dbc5.

bentsku

thanks a lot for removing the changes for the S3 image, sorry it has been a bit complicated with the tests dependencies and had to go further down the path of quick hacks in the make target 😬 but very grateful you've addressed it. Thank you so much!

And sorry again for missing this and coming in late requesting changes. 😞

Great work, can't wait for us to have a unified event ruler! 🚀

andy-g · 2024-04-18T16:03:23Z

@joe4dev I see this was merged, and I see that new docker images have been built, am I right in assuming these changes should be in latest?

I'm running:

docker pull localstack/localstack:latest
latest: Pulling from localstack/localstack
Digest: sha256:e9b46037d6af0b947aad43895edf371b9de96b17cacb52383f5c6c7a2467817a
Status: Image is up to date for localstack/localstack:latest
docker.io/localstack/localstack:latest

Using this rule my lambda works (as it has previously):

{
  "detail-type": ["PaymentReceived","PaymentCompleted"],
  "source": ["producer-1"]
}

...but changing my rule to either of these result in the lambda not being invoked (I was hoping this PR would change that):

{
  "detail-type": ["PaymentReceived"],
  "detail.payload.paymentDetails": {
    "url": [{"exists": true}],
    "type": ["Multi"]
  },
  "source": ["producer-1"]
}

{
  "$or": [
    {
      "detail-type": ["PaymentCompleted"],
      "detail.payload.url": [{"exists": true}]
    },
    {
      "detail-type": ["PaymentReceived"],
      "detail.payload.paymentDetails": {
        "url": [{"exists": true}],
        "type": ["Multi"]
      }
    }
  ],
  "source": ["producer-1"]
}

These are working fine on AWS itself.

Am I missing something here?

andy-g · 2024-04-18T16:09:42Z

Sorry, after a closer look I realised I needed to set the EVENT_RULE_ENGINE environment variable to java. I added that to my docker-compose environment array and it's working like a charm, thanks so much 🍻 .

joe4dev · 2024-04-18T16:17:43Z

I see this was merged, and I see that new docker images have been built, am I right in assuming these changes should be in latest?

Yes, the changes are released for localstack/localstack but NOT yet for localstack/localstack-pro (I just merged a fix for a hidden ARM build issue in our CI, hence the pro image should be released in the next few hours 🤞 ).

@andy-g Awesome to hear everything works as expected using EVENT_RULE_ENGINE=java 🚀 🙌
We wanted to learn more how the Python=>Java bridge works before rolling out that feature by default. That's why we opted for a feature flag. I'm looking forward to hearing your feedback if you notice anything (e.g., performance, stability etc) 👂

joe4dev · 2024-04-18T19:46:42Z

@andy-g I can confirm the localstack/localstack-pro image is now successfully released and the new opt-in event-ruler is working 🚀

joe4dev added the semver: minor Non-breaking changes which can be included in minor releases, but not in patch releases label Apr 8, 2024

joe4dev self-assigned this Apr 8, 2024

joe4dev force-pushed the add-experimental-event-ruler branch from cd43cbd to a67c3c6 Compare April 9, 2024 14:07

joe4dev force-pushed the add-experimental-event-ruler branch from 58a4bd4 to 017bae5 Compare April 11, 2024 21:24

joe4dev commented Apr 12, 2024

View reviewed changes

joe4dev marked this pull request as ready for review April 12, 2024 08:35

joe4dev requested review from dominikschubert, dfangl, pinzon, simonrw, Morijarti, thrau and alexrashed as code owners April 12, 2024 08:35

alexrashed reviewed Apr 12, 2024

View reviewed changes

localstack/services/events/packages.py Outdated Show resolved Hide resolved

localstack/services/events/packages.py Outdated Show resolved Hide resolved

joe4dev force-pushed the add-experimental-event-ruler branch from 0aca7a1 to ec960ea Compare April 15, 2024 11:00

alexrashed reviewed Apr 15, 2024

View reviewed changes

dfangl reviewed Apr 15, 2024

View reviewed changes

joe4dev added 4 commits April 16, 2024 12:17

Add feature flag for Java event ruler implementation

0d89025

Add Java exception handling for event ruler

776c2c7

Add LS package for event ruler libs

a4ea75e

Add comment about machine-based matching

6ecd057

joe4dev force-pushed the add-experimental-event-ruler branch from efaf60c to 9a54628 Compare April 16, 2024 10:22

Switch default back to Python implementation

1ccc192

dfangl approved these changes Apr 17, 2024

View reviewed changes

joe4dev mentioned this pull request Apr 17, 2024

Clarify dev dependencies for host mode localstack/docs#1182

Merged

joe4dev added this to the 3.4 milestone Apr 17, 2024

bentsku reviewed Apr 17, 2024

View reviewed changes

joe4dev added 6 commits April 18, 2024 09:34

Move JPype1 dependency from base-runtime into runtime

1c37ed8

Switch default to java for CI run

4e7dbc5

Fix S3 tests requiring native libs for runtime deps

29f660d

Install JPype runtime deps before installing test deps

cd22f80

Remove unnecessary second g++ install

8f108eb

Revert "Switch default to java for CI run"

dc265aa

This reverts commit 4e7dbc5.

bentsku approved these changes Apr 18, 2024

View reviewed changes

joe4dev merged commit d858f00 into master Apr 18, 2024
41 checks passed

joe4dev deleted the add-experimental-event-ruler branch April 18, 2024 11:01

joe4dev mentioned this pull request Apr 18, 2024

Fix missing config in list for CLI #10689

Merged

joe4dev mentioned this pull request Apr 18, 2024

bug: Lambda Event Source Mapping Filter Criteria not respecting comparison operator exists: false #10325

Closed

1 task

This was referenced Apr 23, 2024

Add event ruler config localstack/docs#1189

Merged

Feat: Eventbridge v2: Add pattern matching #10664

Merged

Add experimental event ruler #10615

Add experimental event ruler #10615

Conversation

joe4dev commented Apr 8, 2024 • edited

Motivation

Changes

Testing

LocalStack Shutdown

Limitations

Kinesisanalytics

TODO

Follow Up

github-actions bot commented Apr 9, 2024 • edited

S3 Image Test Results (AMD64 / ARM64)

github-actions bot commented Apr 11, 2024 • edited

LocalStack Community integration with Pro

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

bentsku Apr 17, 2024 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

bentsku Apr 17, 2024 • edited

Choose a reason for hiding this comment

Choose a reason for hiding this comment

alexrashed left a comment

Choose a reason for hiding this comment

alexrashed left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

joe4dev commented Apr 15, 2024

joe4dev commented Apr 16, 2024

joe4dev commented Apr 17, 2024

joe4dev commented Apr 17, 2024

dfangl left a comment

Choose a reason for hiding this comment

andy-g commented Apr 17, 2024

joe4dev commented Apr 17, 2024

andy-g commented Apr 17, 2024

maxhoheiser commented Apr 17, 2024

bentsku left a comment

Choose a reason for hiding this comment

bentsku left a comment

Choose a reason for hiding this comment

andy-g commented Apr 18, 2024

andy-g commented Apr 18, 2024

joe4dev commented Apr 18, 2024

joe4dev commented Apr 18, 2024

joe4dev commented Apr 8, 2024 •

edited

github-actions bot commented Apr 9, 2024 •

edited

github-actions bot commented Apr 11, 2024 •

edited

bentsku Apr 17, 2024 •

edited

bentsku Apr 17, 2024 •

edited