Failed to get python threadid for main thread when profiling a Dockerized process #168
Comments
|
So - the problem is that we're spamming the log with a non-fatal error message here, and thats getting in the way of the top view =( As a short term fix if you suppress all log messagse with I'll get a proper fix in the next release - |
benfred
added a commit
that referenced
this issue
Sep 27, 2019
When profiling a process inside a docker container from the Host OS, we fail to unwind the native stack (which we use for figuring out the native thread id). This caused an issue with the log being spammed with errors about this obscuring the actual output (#168). Fix by falling back to previous behaviour in this case, and also downgrade the error message to a warning.
benfred
added a commit
that referenced
this issue
Sep 27, 2019
When profiling a process inside a docker container from the Host OS, we fail to unwind the native stack (which we use for figuring out the native thread id). This caused an issue with the log being spammed with errors about this obscuring the actual output (#168). Fix by falling back to previous behaviour in this case, and also downgrade the error message to a warning.
|
fixed in v0.2.1 |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
When profiling a python process from host that is running in a Docker container I'm getting lots of these errors:
Occasionally (like one in 1000 lines) I'm able to see the top view of py-spy for a single sample
This used to work with previous version of py-spy and the issue started occurring with the new 0.2.0 version.
In the container I'm running
Python 2.7.15+and the host is Debian 9.9 running kernel:4.9.0-9-amd64 #1 SMP Debian 4.9.168-1+deb9u5 (2019-08-11) x86_64 GNU/LinuxThe text was updated successfully, but these errors were encountered: