- Adjust logging level of lambda output into OTF logs from debug to info
- Bump devcontainer base image and pre-commit-hooks version
- Alter logging messages for lambda and s3 transfers a little
- Bump workflow action versions
- Add optional min_cache_age kwarg for secrets manager plugin. If specified, multiple requests to update the same secret within this time period will log a warning and not update the variable in secrets manager
- Use boto3 s3 client copy (instead of copy_object) when moving/renaming for post copy actions, to allow handling of large files
- Fix file watch test after text adjustment
- Adjust tests for SSM parameter lookup failure
- Adjust AWS (SSM) lookup plugin to log an error and use "UNKNOWN" for the value of failed secrets, instead of raising exception and breaking all jobs within environment
- Add workaround for botocore issue with socket options for Lambda
- Fix error in JSON schema for Lambda protocol
- Update Lambda protocol to allow overriding default `read_timeout¦ value from 60 seconds. This is needed when executing long running Lambda functions
- Allow blank postCopyAction, or "none". This can be useful when using the same job definition for multiple environments and are using a variable to control the PCA. This way you can set the variable to "none" in the environment where you don't want it to run.
- Also prevent S3 source directories from starting or ending with a /
- Add possibility for encryption to the s3 schemas
- Added new cacheable plugin equivalent to the SSM one, but for Secrets Manager
- Added new cacheable plugin to allow dynamically updated variables to be written back to SSM Parameter Store. For more detail see
open-task-framework
documentation for version 24.22.0 - Minor tweaks to SSM lookup plugin
- Removed a stray
break
in S3 file listing that was preventing fetching more than the first 1000 records - Added new feature to allow moving a file from the source S3 bucket into another destination as a post copy action by referencing the full
s3://
path.
- Added a wider window for refreshing token. If it's within 60 seconds of expiry when checking we will refresh it, so handle instances where there's a delay renewing.
- Fix issue where too many files were being uploaded from worker
- Properly fix STS token refresh code
- Linting updates for ruff
- Added unused argument linting check back in, which would have caught point 1
- Refactor redundant code relating to sessions and STS token refreshing, making the code more generic and moving to
creds.py
instea
black
formatting fix
- Add test coverage for s3 file renames when using proxy transfer type
- Add destination file rename to S3 transfers
- Fix temporary token not refreshing correctly
- Implement required breaking changes from
opentaskpy
v24.13.1
- More logging updates to handle major logging change in
opentaskpy
v24.10.0
- Fix logging for S3 not printing correct destination directory name
- Improve logging for S3 transfers so as to not need DEBUG set to see anything useful
- Fix exceptions thrown my S3 commands etc not causing transfer to fail
- Update lambda and ECS invocations to handle assume role correctly
- Update all AssumeRole usage to also check the expiry of the temporary token and renew them if necessary
- Handle paramstore responses with newlines better
- Ensure that S3 listing only matches the file we want, and not ones in lower directory trees
- Bump version of opentaskpy required
- Alter logging for found files to only print ones that match regex
- Fixes tests that correctly started failing after upgrading opentaskpy version
- Additional checking to PCAs to validate that it has actually happened
- Change version numbering format
- Allow forcing S3 transfers to proxy rather than using CopyObject. This prevents permission issues when trying to do cross account transfers where also using AssumeRole.
- Ensure AssumeRole gets run and creds created correctly
- Add ability to use AssumeRole in protocol definition
- Increase timeout for lambda function creation test
- Add missing schema for S3 transfer - PostCopyAction
- Add ability to trigger Fargate tasks
- Fix minor typo in schemas to use the correct paths in
$id
- Bump required version of
opentaskpy
to v0.13.0
- Add code coverage checks to GitHub workflow
- Fix missing coverage in tests, and related bugs in code
- Update JSON source schemas for S3 transfers.
- Add badges to
README.md
- Add code coverage checks to GitHub workflow
- Updated linting and workflows to match
opentaskpy
repo standards - Updated to work with latest
opentaskpy
version - Fix lots of tests. It seems tests only work when run from an actual machine, and no longer inside a dev container. It seems localstack cannot run lambda functions nested within 2 levels of docker container.
- Added proper JSON validation & add new tests to match
- Altered GitHub workflow to validate existence of
CHANGELOG.md
for each PR - Update requirements for
opentaskpy
to0.6.1
- Added execution handler to call Lambda functions - This allows multiple invocation types, either async or synchronous. Be aware that synchronous execution will block until it's completed. When being used with a batch, the batch cannot kill the running lambda function if it times out before the function is completed.