Releases: ebarault/anyservice-awscode-docker
Releases · ebarault/anyservice-awscode-docker
Improve performance of log events printing + bug fixes
Fix: iterate on same token while there are no new log events
3.5.6 fix: iterate on same token while there are no new log events
bug fix
bug fix
Upload Artefacts for all builds | Customize gitlab artifacts dir
bug fixes
Fetch codebuild artifacts from S3 bucket
Fix case with no artifacts.zip available (#3) * get codebuild to gitlab-runner * get codebuild to gitlab-runner * get codebuild artifacts to runner * python3 * wip * wip * wip * wip * README.md * README.md * support ARTIFACTS_PACKAGING NONE | ZIP * support ARTIFACTS_PACKAGING NONE | ZIP * support ARTIFACTS_PACKAGING NONE | ZIP * support ARTIFACTS_PACKAGING NONE | ZIP * support ARTIFACTS_PACKAGING NONE | ZIP * support ARTIFACTS_PACKAGING NONE | ZIP * support ARTIFACTS_PACKAGING NONE | ZIP * fix doc * fix doc * Catch error if job does not have artifact * fix if * fix if * fix merge
improve fetch_logs routine
The fetch_logs routine currently iterates every 10s to fetch new logs, displaying on all logs returned by a single call to aws logs get-log-events.
If there are more logs available (+1Mb / +10,000 events) they are not fetched until the next iteration, wasting 10s each time, with the impact that the build time at gitlab-ci is longer than the real build time, and some logs are never fetched.
AWS reference:
https://docs.aws.amazon.com/cli/latest/reference/logs/get-log-events.html
- aws logs get-log-events fetches as many log events as can fit in a response size of 1 MB, up to 10,000 log events. it returns nextForwardToken, a token for the next set of items in the forward direction. If you have reached the end of the stream, it will return the same token you passed in.