slow keep-alive with generic kernel

classic Classic list List threaded Threaded
2 messages Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

slow keep-alive with generic kernel

S.A.N
There are 2 different versions of the program, using keep-alive
First program, between the call to the connect operation and the write
operation, there is a short amount of time
Second program, write operation is called immediately after the connect
operation.
At the same time, the first program shows a rps of 2 times less than the
second program.
After tracing tcpdump, I can see that the server seems to "remember" the
time interval between the connection and the first write, and then follows
it when sending all subsequent responses.
With what it can be connected?

Repeats on linux kernel 4.4.0-31-generic

Posted at Nginx Forum: https://forum.nginx.org/read.php?2,273276,273276#msg-273276

_______________________________________________
nginx mailing list
[hidden email]
http://mailman.nginx.org/mailman/listinfo/nginx
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

Re: slow keep-alive with generic kernel

S.A.N
Here is a demo example
https://gist.github.com/magisterRab/6b7132e0b9e88baa4b7e0e69a2ff0aab
if in line 120 remove writeRequest(fd), then the speed of the test will fall
2 times

Posted at Nginx Forum: https://forum.nginx.org/read.php?2,273276,273308#msg-273308

_______________________________________________
nginx mailing list
[hidden email]
http://mailman.nginx.org/mailman/listinfo/nginx
Loading...