HI All, Is Dagster Serverless down ? Could not re...
# dagster-plus
j
HI All, Is Dagster Serverless down ? Could not reach user code server. gRPC Error code: UNKNOWN
p
Hi, do you mind sharing your org name? I can take a look…
j
@prha Oath
@prha The other error which happens which I can reach the server is URL
Copy code
dagster_dbt.errors.DagsterDbtCliOutputsNotFoundError: Expected to find file at path /root/.pex/user_code/8eb924ed1df63bf8256e21667c4e9aad1c9a7683/working_directory/root/assets_dbt_python/../dbt_project/target/run_results.json
p
I see the serverless agent and code locations up for your project… may have been some temporary error that we’ve recovered from, though I can see the same error in the logs
Can you successfully deploy your code location?
I’m not sure about those run errors… may need to consult with some other folks on the team
j
Yes, I can deploy succesfully
@prha
👍 1
@prha Job run fails and has suddenly started failing. I know it's not a code issue on my end. Could you connect me with someone ?
p
So, the latest theory is that running the job is causing OOM issues on the code location, which causes it to restart.
When you kick off your job, you might have better luck checking the
Isolate Run Environment
option, which should be available as an option in the configuration editor
This will increase the start-up time of your runs, but should prevent runs from affecting the code location
j
Thanks. Trying that
p
For the dbt error in the run logs, are you by any chance changing the target path in your
dbt_project.yml
file? This is something that we can’t handle very well (reference: https://github.com/dagster-io/dagster/issues/10918).
j
I read that; i did not change anything. Job just stopped working suddenly. It's working on my laptop.
d
@jitesh soni what version of dbt-core do you have installed on your laptop? A new version of that just came out today
j
Hi @prha , I am unable to find this
Isolate Run Environment
option . Screenshot attached
p
You’ll probably have to hold down shift and click the materialize button in order to see those options.
j
Do not see it 😞
d
Ah shift-click on the “materialize all” text not the drop down
j
Server not found 😞
p
I disabled non-isolated runs for your deployment, so you might be able to try again. For the dbt error, I wonder if you could upgrade your
dbt-core
package locally and repro the error? If it does repro locally with that version (which released today), you might want to pin the package to an earlier version.
The isolated run setting can be found by clicking the deployment selector (the gear next to
prod
in the dropdown) and flip the value:
Copy code
non_isolated_runs:
  enabled: false
j
Hi @prha , Host are still not reachable!
Copy code
The above exception was caused by the following exception:
grpc._channel._InactiveRpcError: <_InactiveRpcError of RPC that terminated with:
status = StatusCode.UNKNOWN
details = "Exception calling application: No server created with the given handle"
debug_error_string = "{"created":"@1674794708.234532664","description":"Error received from peer ipv4:10.0.94.163:4000","file":"src/core/lib/surface/call.cc","file_line":966,"grpc_message":"Exception calling application: No server created with the given handle","grpc_status":2}"
>
@prha Hold it might have worked. I was using Brave Browser and I had to disable Brave Shield!
🎉 1
@prha Thank you for all your help and patience !!!! I can confirm that after disabling Brave Shield on my Brave Browser; host not reachable issues went away.
p
I’m not sure how the browser would have any effect on your grpc server, but I am glad that you are unblocked now.
j
Folks, I checked the same issue occurred again on multiple days. Any suggestions ?
Copy code
FileNotFoundError: [Errno 2] No such file or directory: '/root/.pex/user_code/b64440d5603ca80da76721e1a6eaa442b73252b7/working_directory/root/assets_dbt_python/../dbt_project/target/run_results.json'
  File "/root/.pex/installed_wheels/77b42134a058a50fb83cd3926911571c40f8a31dfcf3a9f3584f5e74ebb969c7/dagster_dbt-0.17.13-py3-none-any.whl/dagster_dbt/cli/utils.py", line 168, in parse_run_results
    with open(run_results_path, encoding="utf8") as file:
The above exception occurred during handling of the following exception:
UnicodeDecodeError: 'utf-8' codec can't decode byte 0x99 in position 16: invalid start byte
d
Do you possibly have dbt-core 1.4.0 installed? The latest version of dagster added a temporary pin to dbt-core 1.3.0 until we can fix some breaking changes in the dbt-core release
so you could either rebuild with the latest dagster version or pin to dbt-core 1.3.0
j
@daniel I have not upgraded local to 1.4.0 . What's confusing to me is that I do not supply dbt when I deploy to Dagster. Where do I specify which version of DBT is deployed to Dagster ?
d
dagster-dbt depends on dbt-core - i think the easiest way to resolve this would be to upgrade to the latest version of dagster (1.1.14) where we have a pin applied to fix a breaking change introduced in dbt-core 1.4.0
s
Hello - I looked into it briefly as well. It appears the project is using dbt-core 1.4.1 and dagster 1.1.13. This just depends on when the project was first deployed from github. As daniel suggested, setting
dagster==1.1.14
in your setup.py should force a rebuild of your dependencies and pull in the pinned dbt-core.
👍 1