Workflow and playwright test updates#1360
Conversation
2cd9ac3 to
30d7f76
Compare
| HealthChecksUI__HealthChecks__9__Uri: 'http://ordering-signalrhub:5112/liveness' | ||
| HealthChecksUI__HealthChecks__10__Name: 'Ordering HTTP Background Check' | ||
| HealthChecksUI__HealthChecks__10__Uri: 'http://ordering-backgroundtasks:5111/hc' | ||
| HealthChecksUI__HealthChecks__10__Uri: 'http://ordering-backgroundtasks:5111/liveness' |
There was a problem hiding this comment.
were we using the health check incorrectly before?
There was a problem hiding this comment.
The eshop code has a readiness check at /hc and a liveness check at /liveness. We had the readiness check hooked up as a liveness check
| run: | | ||
| if [[ "${{ matrix.container }}" != "" ]]; then | ||
| rad resource expose containers ${{ matrix.container }} ${{ matrix.exposeArgs }} --port ${{ matrix.port }} & | ||
| export ENDPOINT="http://localhost:3000/" |
There was a problem hiding this comment.
| export ENDPOINT="http://localhost:3000/" | |
| export ENDPOINT="http://localhost:${{ matrix.port }}" |
090f61b to
2f36308
Compare
| - edge | ||
| schedule: # Run every 2 hours | ||
| - cron: "0 */2 * * *" | ||
| concurrency: |
There was a problem hiding this comment.
Can you add a comment explaining why this is needed? It's easy to forget the rationale for adding something like this later.
There was a problem hiding this comment.
This is to group workflow runs and cancel if a new workflow starts running. What I mean is that, if we create a PR, a workflow is triggered. Then if we push another commit to that PR, another workflow will be triggered and the in-progress one should be cancelled.
Now that I think about this, because we might have already created resources, I think it would be good to not cancel this workflow and make sure that the resources are deleted if created.
| if: steps.gen-id.outputs.RUN_TEST == 'true' | ||
| id: deploy-app | ||
| run: rad deploy ${{ matrix.path }} ${{ matrix.deployArgs }} -e ${{ matrix.env }} | ||
| - name: Wait for all pods to be ready |
There was a problem hiding this comment.
Do we no longer care about waiting for the pods to be ready?
There was a problem hiding this comment.
I think we discussed this in one of our 1:1 meeting. rad deploy should wait for all pods to be ready before returning successfully. Is that wrong?
|
Is this working better now that you swapped the health checks? |
2f36308 to
891c135
Compare
Yes, it definitely is working better. Not passing the tests all the time but still a good improvement so far with the changes here. |
8c34b1a to
6e808c4
Compare
6e808c4 to
3c3b66c
Compare
Signed-off-by: ytimocin <ytimocin@microsoft.com>
3c3b66c to
cbeaf5e
Compare
| # Empty the file | ||
| truncate -s 0 $DELETED_RESOURCES_FILE | ||
|
|
||
| for resource_type in ${RESOURCE_TYPES//,/ } |
There was a problem hiding this comment.
why the changes to this file?
Signed-off-by: ytimocin <ytimocin@microsoft.com> Signed-off-by: sk593 <shruthikumar@microsoft.com>
Signed-off-by: ytimocin <ytimocin@microsoft.com> Signed-off-by: sk593 <shruthikumar@microsoft.com>
* update gh to ado sync workflow to use service principals (#1363) * Make env configurable (#1046) Signed-off-by: Reshma Abdul Rahim <reshmarahim.abdul@microsoft.com> * Adding test AKS workflow to v0.32 (#1052) * Run functional tests every 2 hours and add another workflow that runs… (#1020) * Run functional tests every 2 hours and add another workflow that runs the tests on AKS instead of k3d Signed-off-by: ytimocin <ytimocin@microsoft.com> * Triggering workflow Signed-off-by: ytimocin <ytimocin@microsoft.com> --------- Signed-off-by: ytimocin <ytimocin@microsoft.com> * Removing the run of the AKS workflow on PRs (#1051) Signed-off-by: ytimocin <ytimocin@microsoft.com> * Update the timeout (30s) for Playwright Signed-off-by: ytimocin <ytimocin@microsoft.com> --------- Signed-off-by: ytimocin <ytimocin@microsoft.com> * Use retry and update eshop playwright tests to wait for the catalog to appear (#1213) Signed-off-by: ytimocin <ytimocin@microsoft.com> * update gh to ado sync workflow to use service principals Signed-off-by: Will Tsai <28876888+willtsai@users.noreply.github.com> --------- Signed-off-by: Reshma Abdul Rahim <reshmarahim.abdul@microsoft.com> Signed-off-by: ytimocin <ytimocin@microsoft.com> Signed-off-by: Will Tsai <28876888+willtsai@users.noreply.github.com> Signed-off-by: Will <28876888+willtsai@users.noreply.github.com> Co-authored-by: Reshma Abdul Rahim <61033581+Reshrahim@users.noreply.github.com> Co-authored-by: Yetkin Timocin <ytimocin@microsoft.com> Signed-off-by: sk593 <shruthikumar@microsoft.com> * Workflow and playwright test updates (#1360) Signed-off-by: ytimocin <ytimocin@microsoft.com> Signed-off-by: sk593 <shruthikumar@microsoft.com> * Changing the place of page.reload in the tests (#1492) Signed-off-by: ytimocin <ytimocin@microsoft.com> Signed-off-by: sk593 <shruthikumar@microsoft.com> * Adding a job that will purge AWS EKS clusters every 6 hours (#1462) Signed-off-by: ytimocin <ytimocin@microsoft.com> Signed-off-by: sk593 <shruthikumar@microsoft.com> * Update all deps of all apps (#1536) Signed-off-by: ytimocin <ytimocin@microsoft.com> Signed-off-by: sk593 <shruthikumar@microsoft.com> * Updating the runner of purge azure test resources workflow (#1537) Signed-off-by: ytimocin <ytimocin@microsoft.com> Signed-off-by: sk593 <shruthikumar@microsoft.com> * Update purget test resources to fix the workflow failures (#1552) Signed-off-by: ytimocin <ytimocin@microsoft.com> Signed-off-by: sk593 <shruthikumar@microsoft.com> * Fixing the eShop catalog load issue for the UI tests (#1548) Signed-off-by: ytimocin <ytimocin@microsoft.com> Signed-off-by: sk593 <shruthikumar@microsoft.com> --------- Signed-off-by: Reshma Abdul Rahim <reshmarahim.abdul@microsoft.com> Signed-off-by: ytimocin <ytimocin@microsoft.com> Signed-off-by: Will Tsai <28876888+willtsai@users.noreply.github.com> Signed-off-by: Will <28876888+willtsai@users.noreply.github.com> Signed-off-by: sk593 <shruthikumar@microsoft.com> Co-authored-by: Will <28876888+willtsai@users.noreply.github.com> Co-authored-by: Reshma Abdul Rahim <61033581+Reshrahim@users.noreply.github.com> Co-authored-by: Yetkin Timocin <ytimocin@microsoft.com>
Wait for all pods to be readystep because the previous step,rad deploy, should do the waiting already.Cancel Workflowis called.Delete EKS Clusterstep. (Write EKS cluster name to issue title or description #1003)rad uninstallfrom theDelete EKS Clusterstep. (EKS cluster not deleted if rad not installed #868)Fixes:
#1003
#868
#1017