When adding a Dagster Cloud location if I pass ` module name

Question

When adding a Dagster Cloud location if I pass ` module name` the CLI times out waiting for the update and the UI just says Loading for hours I don t see an error from the agent The <https docs dagster cloud guides adding code|code location docs> say to use ` package name` and don t mention ` module name` Would it make sense to remove the ` module name` option from the `dagster cloud` CLI entirely It would be nice to either get an error message or not have the option to do it the wrong way

daniel · Answer

Hi Mark module name and package name should be identical To confirm it s working for you when you set package name but not module name

daniel · Answer

Do you have a link to the location in cloud that you could share here or via DM

daniel · Answer

should behave identically that is

Mark Fickett · Answer

Correct worked with package name but when I previously tired module name it wouldn t load same ECR image It s the `tech data pipeline ecr` location in <https formenergy dagster cloud prod workspace>

Mark Fickett · Answer

I just created `tech data pipeline ecr with module name for debug` with the same location definition except back to `module name`

daniel · Answer

Do you know what version of the dagster cloud library you re running

Mark Fickett · Answer

```$ pip list | grep dagster cloud dagster cloud 0 14 15```

daniel · Answer

What about in the agent itself

Mark Fickett · Answer

Is the agent hosted by Dagster Cloud Or how do I check

daniel · Answer

You re hosting the agent the answer depends on whether you re in K8s ECS etc

Mark Fickett · Answer

Ah Kubernetes hosted by AWS

Mark Fickett · Answer

I m not very practiced with k8s so I may be using the wrong terminology here

daniel · Answer

No problem I believe we added support for module name in a version of the agent that s later than what you have but I m surprised that you re not getting an error back from the Dagster Cloud CLI when you try to use module name we ll look into that In the meantime though I suspect running a helm upgrade of the agent in your k8s cluster will allow you to use module name similar to the instructions here at step 4 <https docs dagster cloud agents kubernetes setup>

Mark Fickett · Answer

I vaguely remember starting the agent but haven t updated it since 2021 fall at least Here s the status When I view metadata it says ``` type null version null ```

daniel · Answer

Once you upgrade I suspect that version will populate as well

Mark Fickett · Answer

I attempted to upgrade ```$ helm repo update $ helm upgrade gt install user cloud dagster cloud dagster cloud agent gt namespace dagster cloud gt set dagsterCloud deployment=prod Release user cloud has been upgraded Happy Helming NAME user cloud LAST DEPLOYED Wed Jun 1 10 13 51 2022 NAMESPACE dagster cloud STATUS deployed REVISION 8 TEST SUITE None``` and I see ```$ helm list namespace dagster cloud NAME NAMESPACE REVISION UPDATED STATUS CHART APP VERSION user cloud dagster cloud 8 2022 06 01 10 13 51 449548477 0400 EDTdeployed dagster cloud agent 0 14 17 0 14 17``` but I m not seeing any active agent old one became inactive new one has not appeared after 30m grimacing

daniel · Answer

60 seconds should be plenty for it to appear Any logs from the agent pod

daniel · Answer

My best guess would be needing to set the agent token secret step 3 here <https docs dagster cloud agents kubernetes setup> it s possible back when you set this up late last year it was done slightly differnetly and you included the secret in the helm upgrade command

Mark Fickett · Answer

Still looking for the logs Seems like the secret already existed ```$ kubectl create secret generic dagster cloud agent token from literal=DAGSTER CLOUD AGENT TOKEN=agent lt mytoken gt namespace=dagster cloud error failed to create secret secrets dagster cloud agent token already exists $ kubectl get secret namespace dagster cloud NAME TYPE DATA AGE dagster cloud agent token Opaque 1 216d``` However I deleted and re created it ```$ kubectl get secret namespace dagster cloud NAME TYPE DATA AGE dagster cloud agent token Opaque 1 54s```

daniel · Answer

Got it try kubectl get pods and then hopefully there s a pod with agent in the name if its not running kubectl logs lt that pod ID gt will hopefully have some clues or kubectl describe pod lt that pod ID gt

Mark Fickett · Answer

```user cloud dagster cloud agent agent 5689bb4d6c bc92s 0 1 CrashLoopBackOff 12 40m```

Mark Fickett · Answer

```$ kubectl logs user cloud dagster cloud agent agent 5689bb4d6c bc92s namespace dagster cloud Traceback most recent call last File usr local bin dagster cloud line 33 in lt module gt sys exit load entry point dagster cloud console scripts dagster cloud File usr local lib python3 8 site packages typer main py line 214 in call return get command self args kwargs File usr local lib python3 8 site packages click core py line 1130 in call return self main args kwargs File usr local lib python3 8 site packages click core py line 1055 in main rv = self invoke ctx File usr local lib python3 8 site packages click core py line 1657 in invoke return process result sub ctx command invoke sub ctx File usr local lib python3 8 site packages click core py line 1657 in invoke return process result sub ctx command invoke sub ctx File usr local lib python3 8 site packages click core py line 1404 in invoke return ctx invoke self callback ctx params File usr local lib python3 8 site packages click core py line 760 in invoke return callback args kwargs File usr local lib python3 8 site packages typer main py line 500 in wrapper return callback use params type ignore File dagster cloud dagster cloud agent cli init py line 179 in run run local agent in environment dagster home File dagster cloud dagster cloud agent cli init py line 57 in run local agent in environment run local agent File dagster cloud dagster cloud agent cli init py line 40 in run local agent with DagsterCloudAgentInstance get as instance File dagster cloud dagster cloud instance init py line 200 in get instance = DagsterInstance get File dagster dagster core instance init py line 409 in get return DagsterInstance from config dagster home path File dagster dagster core instance init py line 424 in from config return DagsterInstance from ref instance ref File dagster dagster core instance init py line 436 in from ref return klass type ignore File dagster cloud dagster cloud instance init py line 60 in init assert self dagster cloud url File dagster cloud dagster cloud instance init py line 111 in dagster cloud url raise DagsterInvariantViolationError dagster core errors DagsterInvariantViolationError Could not derive Dagster Cloud URL from agent token Create a new agent token or set the `url` field under `dagster cloud api` in your `dagster yaml` ```

Mark Fickett · Answer

I see if I generate a new agent token the format is different I ll reset that

daniel · Answer

Cool two options here Re run the helm upgrade command with an additional ``` set dagsterCloud organization= lt name of your organization gt ``` or yeah generate a new token

daniel · Answer

either will work the latter will make it not happen again in the future

Mark Fickett · Answer

Great after updating the agent token the pod retried and was able to run so I ve got an updated agent going Thanks And it does show metadata ``` type K8sUserCodeLauncher version 0 14 17 ```