Hi, I am running a Kubernetes deployment of Dagster, but I cannot see raw compute logs (stdout / stderr). I saw in helm’s

values.yaml

file that this stuff is handled by

computeLogManager

. Should I switch from

LocalComputeLogManager

to another one? perhaps

S3ComputeLogManager

? Or can I do something to make it work with ``LocalComputeLogManager`` itself?

Should I switch from LocalComputeLogManager  to another one? perhaps S3ComputeLogManager ?

Yea thats likely the right move.

Thanks, just mentioning here that I am using a

k8sRunLauncher

for this, hope it will work with that

cc @johann

You can create the secret in the

extraManifests:

field of your helm values, then reference it in the config of your run launcher, with

env_secrets

so that launched pipelines will have the secret attached.

thanks, I already have

envSecrets

under

k8sRunLauncher

. Should that be fine? Or do I have to take the long route of

extraManifests

?

envSecrets specifies secrets to attach to your pods. It won’t create the k8s secret object for you

oh, yeah, so I create the secret using

kubectl apply -f xxx.yaml

before

Gotcha. In that case you don’t need to use extraManifests

thanks a lot, I will give this a try

I believe it will be created, but could you try it and report back?

okay sure

likely you need to update the

dagit

deployment section to include the secrets since its whats attempting to read

aah I see