Open tituslhy opened 1 day ago
The issue you're experiencing with the WorkflowTimeoutError
in your nested workflow setup on LlamaDeploy seems to be related to the default timeout setting. The error message indicates that the timeout is occurring after 10.0 seconds, which is the default timeout for workflows in LlamaDeploy [1].
Here are a few things to check and try:
Ensure Timeout Override: Make sure that the timeout settings you specified for both the InnerWorkflow
and OuterWorkflow
are correctly applied. It seems like the default timeout of 10.0 seconds is being used instead of your specified timeouts. Double-check that the timeout
parameter is being correctly passed and applied when you instantiate your workflows.
Check add_workflows
Method: When you add the InnerWorkflow
to the OuterWorkflow
using outer.add_workflows(inner=InnerWorkflow())
, ensure that the instance of InnerWorkflow
being added has the correct timeout setting. You might want to pass the already configured inner
instance instead of creating a new one with InnerWorkflow()
.
Deployment Configuration: Verify that the deployment configuration for both workflows is correct and that there are no issues with how the workflows are being deployed or how the AsyncLlamaDeployClient
is interacting with them.
Debugging: Add logging or print statements to confirm that the timeout values are being set as expected when the workflows are initialized and deployed.
By ensuring that the correct timeout values are being applied and that the InnerWorkflow
is properly configured when added to the OuterWorkflow
, you should be able to resolve the timeout issue.
To continue talking to Dosu, mention @dosu.
@dosu Sorry I made a typo. This was how I added the inner workflow
inner = InnerWorkflow(timeout=1200.0,verbose=True)
outer = OuterWorkflow(timeout=6000.0, verbose = True)
outer.add_workflows(inner=inner)
I went to the raw file of LlamaIndex's workflow.py
on my machine and changed the timeout settings. It works now. This means that for some reason, the nested workflow is still only calling on a default setting instead of registering my provided timeout setting.
Hey @tituslhy! It looks like you're sharing some code details, but I'm just a bot here to help with bugs, technical questions, and engineering knowledge. If you think I should have more to say on this, feel free to reach out to the Dosu Team. They'll know what to do! 😊
To continue talking to Dosu, mention @dosu.
Hey @tituslhy thanks for reporting the issue, I'll try to reproduce and will update the issue with my findings.
Bug Description
Despite specifying a generous timeout limit for all my workflows, my inner workflow of my nested workflow always times out within 10 seconds once deployed on LlamaDeploy.
Which means that the timeout error only happens when I ping the outer workflow which then routes to the inner workflow.
Version
0.11.17
Steps to Reproduce
Deployment of core
Deployment of workflow
Calling of workflow
Relevant Logs/Tracbacks