Some people blamed me this is incorrect test. I repeated the test with well
known and well coded server APACHE 1.3.9 via 10Mbps local network. I have
runned 50 process each trying to connect a server in a loop.
The server and network didn't have much load. Here is the output of VMSTAT
for the server running this test:
1 0 0 46968 8564 1864 26180 0 0 0 1 1060 54 4 12
84
5 0 0 46968 8852 1864 26188 0 0 1 2 888 85 8 13
79
0 0 0 46968 9660 1864 26188 0 0 0 1 826 51 3 11
86
2 0 0 46968 9996 1864 26192 0 0 0 0 841 82 7 12
81
3 0 0 46968 10332 1864 26192 0 0 0 1 961 70 5 11
84
4 0 0 46968 10852 1864 26192 0 0 0 0 917 63 4 12
83
0 0 0 46968 11984 1868 26192 0 0 0 1 849 62 3 11
86
0 0 0 46968 13920 1868 26192 0 0 0 2 825 61 4 11
85
Here is the concurently going ping from the test machine to a server
machine:
64 bytes from 212.24.53.193: icmp_seq=73 ttl=255 time=0.4 ms
64 bytes from 212.24.53.193: icmp_seq=74 ttl=255 time=0.4 ms
64 bytes from 212.24.53.193: icmp_seq=75 ttl=255 time=18.3 ms
64 bytes from 212.24.53.193: icmp_seq=76 ttl=255 time=0.4 ms
64 bytes from 212.24.53.193: icmp_seq=77 ttl=255 time=0.4 ms
64 bytes from 212.24.53.193: icmp_seq=78 ttl=255 time=24.6 ms
64 bytes from 212.24.53.193: icmp_seq=79 ttl=255 time=0.4 ms
64 bytes from 212.24.53.193: icmp_seq=80 ttl=255 time=0.4 ms
64 bytes from 212.24.53.193: icmp_seq=81 ttl=255 time=0.4 ms
64 bytes from 212.24.53.193: icmp_seq=82 ttl=255 time=0.8 ms
64 bytes from 212.24.53.193: icmp_seq=83 ttl=255 time=0.4 ms
64 bytes from 212.24.53.193: icmp_seq=84 ttl=255 time=0.4 ms
64 bytes from 212.24.53.193: icmp_seq=85 ttl=255 time=0.4 ms
So this is no network load problem or server load problem. Still Some of
connect takes such long as 3,9,21 or 45 second. No other times found. This
is strange.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Fri Apr 07 2000 - 21:00:18 EST