Find cause of status code 0

[grillorafael]

Hi All,

I'm trying to debug why sometimes my requests fail (Doing some perf test through locust.io) but I can't find the cause. My nginx logs have no error attached and what I see is requests giving status code 0.

I thought it could be timeout but I timed the request and I see <1sec.

Is there a way for me to find out why I'm getting this behavior?

I'm basically getting status_code==0 but I can't find a reason to that.

I'm using python 2.7 and locust to run the perf test

[sigmavirus24]

I believe we pull the status code here which means fundamentally this is a urllib3 bug. Requests defaults the status code to None but has specific handling for this case. That aside, is this a non-exceptional case? I've never seen a status code of 0 for a non-exceptional case.

[grillorafael]

I see. And do you know any way for me to debug the reason why urllib3 is failing to do something?

[sigmavirus24]

Looking at response.raw would be a good place to start. You didn't really answer my question though: Is the response object you're receiving coming from an exception that's been handled?

[grillorafael]

I believe so.

This is the line:
https://github.com/locustio/locust/blob/a4c377efc7914e8d90701d9226a5cd843859b790/locust/clients.py#L161

[sigmavirus24]

Did you notice https://github.com/locustio/locust/blob/a4c377efc7914e8d90701d9226a5cd843859b790/locust/clients.py#L164 ?

[Lukasa]

Closing as a downstream bug.