[Dashboard] Getting stacktrace from the dashboard triggers "failed to get os threadid" when --native option is used #30566
Description
What happened + What you expected to happen
When I click the stack trace for the JobSupervisorActor, I saw the following error.
Failed to execute `sudo -n $(which py-spy) dump -p 830 --native`.
Note that this command requires `py-spy` to be installed with root permissions. You
can install `py-spy` and give it root permissions as follows:
$ pip install py-spy
$ sudo chown root:root `which py-spy`
$ sudo chmod u+s `which py-spy`
Alternatively, you can start Ray with passwordless sudo / root permissions.
=== stdout ===
Process 830: ray::JobSupervisor
Python v3.7.7 (/home/ray/anaconda3/bin/python3.7)
=== stderr ===
Error: failed to get os threadid
Even when I run the same command from the terminal, it failed with the same error
Based on the search, I found only one issue, benfred/py-spy#490, but this seems irrelevant to our case (it says it should work if this command runs inside a container).
We may need to figure out the root cause or remove --native as a default option (or we can allow users to turn on/off the flag).
Versions / Dependencies
master
Reproduction script
not sure how to reproduce, but it is easy to find this problem from nightly test.
Issue Severity
No response