I've coded the backend of my recently launched mobile application using django 1.9, and the database I've used is MySQL. I'm about to start a large promotional campaign so trying to gear up the backend for the same.
One issue that I'm constantly facing is that when I fire just around 100 threads in a span of 10 seconds via Apache Jmeter, a large number (~60) of them do not get served. Django server logs look like this:
[24/Aug/2016 22:43:01] "GET /my/api/ HTTP/1.0" 200 78547
[24/Aug/2016 22:43:01] "GET /my/api/ HTTP/1.0" 200 78547
- Broken pipe from ('127.0.0.1', 53582)
- Broken pipe from ('127.0.0.1', 53597)
[24/Aug/2016 22:43:01] "GET /my/api/ HTTP/1.0" 200 78547
- Broken pipe from ('127.0.0.1', 53551)
What exactly does the broken pipe signify? How do I get around this?