Messages in this thread | | | Date | Mon, 24 Dec 2001 18:15:02 +0900 | From | Choi Dong Joon <> | Subject | [Q] congestion window of sender |
| |
I am measuring TCP performance with long fat network (e.g. satellite network) - 540 msec RTT and 155Mbps bandwidth (ATM, CLIP).
9 Mbytes window size for the maximum throughput.
So I changed the tcp window size defined in the following proc files.
/proc/sys/net/core/rmem_default /proc/sys/net/core/rmem_max /proc/sys/net/core/wmem_default /proc/sys/net/core/rmem_max
I measured ttcp or iperf and analyzed it using tcptrace after tcpdump.
Sometime (usually after reboot) I could get maximum throughput but many times I could not get reasonable throughput. When I tcpdump it, analyzed by tcptrace and xplot, the tcp congestion window size of sender does not increase even though the advertized window by receiver is enough large.
But after long time without anything, the TCP works normally. (I could get the maximum throughput)
What is the problem ?
I use redhat 7.1 and kernel 2.4.16.
- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
| |