My workflow has failed multiple times when invoking ct.sh with the below error. My guess is that the chart-testing artifact retrieval fails and is incomplete so the subsequent untar fails. Of course, nothing this action can do if GitHub artifacts timeout but I think we can build in some resiliency to the script. Can we perhaps add --retry 5 --retry-delay 1 to the curl command?
Run helm/chart-testing-action@v2.3.1
with:
version: v3.7.1
yamllint_version: 1.[2](https://github.com/hashgraph/hedera-mirror-node/actions/runs/4287672375/jobs/7469131183#step:7:2)7.1
yamale_version: [3](https://github.com/hashgraph/hedera-mirror-node/actions/runs/4287672375/jobs/7469131183#step:7:3).0.[4](https://github.com/hashgraph/hedera-mirror-node/actions/runs/4287672375/jobs/7469131183#step:7:4)
env:
JAVA_HOME: /opt/hostedtoolcache/Java_Temurin-Hotspot_jdk/17.0.[6](https://github.com/hashgraph/hedera-mirror-node/actions/runs/4287672375/jobs/7469131183#step:7:6)-10/x64
JAVA_HOME_17_X64: /opt/hostedtoolcache/Java_Temurin-Hotspot_jdk/17.0.6-10/x64
GRADLE_BUILD_ACTION_SETUP_COMPLETED: true
GRADLE_BUILD_ACTION_CACHE_RESTORED: true
Run cd $GITHUB_ACTION_PATH \
cd $GITHUB_ACTION_PATH \
&& ./ct.sh \
--version v3.7.1 \
--yamllint-version 1.27.1 \
--yamale-version 3.0.4
shell: /usr/bin/bash --noprofile --norc -e -o pipefail {0}
env:
JAVA_HOME: /opt/hostedtoolcache/Java_Temurin-Hotspot_jdk/1[7](https://github.com/hashgraph/hedera-mirror-node/actions/runs/4287672375/jobs/7469131183#step:7:7).0.6-10/x64
JAVA_HOME_17_X64: /opt/hostedtoolcache/Java_Temurin-Hotspot_jdk/17.0.6-10/x64
GRADLE_BUILD_ACTION_SETUP_COMPLETED: true
GRADLE_BUILD_ACTION_CACHE_RESTORED: true
Installing chart-testing...
gzip: stdin: not in gzip format
tar: Child returned status 1
tar: Error is not recoverable: exiting now
Error: Process completed with exit code 2.
My workflow has failed multiple times when invoking ct.sh with the below error. My guess is that the chart-testing artifact retrieval fails and is incomplete so the subsequent untar fails. Of course, nothing this action can do if GitHub artifacts timeout but I think we can build in some resiliency to the script. Can we perhaps add
--retry 5 --retry-delay 1
to the curl command?