[tor-bugs] #33211 [Circumvention/Snowflake]: proxy-go sometimes gets into a 100+% CPU state

[Tor Bug Tracker & Wiki]

#33211: proxy-go sometimes gets into a 100+% CPU state
-------------------------------------+-----------------------------------
 Reporter:  dcf                      |          Owner:  (none)
     Type:  defect                   |         Status:  needs_information
 Priority:  Medium                   |      Milestone:
Component:  Circumvention/Snowflake  |        Version:
 Severity:  Normal                   |     Resolution:
 Keywords:                           |  Actual Points:
Parent ID:                           |         Points:
 Reviewer:                           |        Sponsor:
-------------------------------------+-----------------------------------

Comment (by dcf):

 Replying to [comment:2 cohosh]:
 > Replying to [comment:1 dcf]:
 > > In the short term, I'd like to disable proxy-go-restartless, so we
 don't have a process that gets permanently wedged. The other proxy-go
 instances presumably also encounter the same problem, but they restart
 themselves every so often.
 > >
 > > Ok?
 > Sounds good to me. Thanks for catching this.

 Okay, I ran `sv down /etc/service/snowflake-proxy-restartless` at
 2020-02-10 22:51.

 Replying to [comment:3 cohosh]:
 > The patch also looks fine, though note that the function getToken is now
 unused.
 >
 > We could just deploy a version of the proxy-go with the profiling code
 without merging to master. I'd prefer that unless we feel the need to do
 profiling longer term.

 Sorry, I didn't mean to propose merging or deploying that patch. I think
 it's sufficient to try it locally. I was able to reproduce just by running
 proxy-go and restarting the client every few hours, but I wasn't profiling
 at the time.

--
Ticket URL: <https://trac.torproject.org/projects/tor/ticket/33211#comment:4>
Tor Bug Tracker & Wiki <https://trac.torproject.org/>
The Tor Project: anonymity online