Throughput deadlock

Edwin Law (e-law@cs.yorku.ca)
Thu, 2 Jul 1998 20:41:09 -0400 (EDT)

Messages sorted by: [ date ][ thread ][ subject ][ author ]
Next message: Anne Wong: "HK FINE WINE SHOP ON NET"
Previous message: Rick Jones: "unsubscribing"
In reply to: Rick Jones: "Re: Different throughputs in two directions"
Next in thread: Rick Jones: "Re: Throughput deadlock"

Hi,
Has anyone tried using a send size of 1 KB and socket size of 4KB in
an Netperf TCP_STREAM test? I got an exceptionally low throughput of
0.04Mbps. I believe it is caused by the so-called "throughput deadlock"
problem, since 4KB < 3MSS and that turning on the TCP_NODELAY can
eliminate this anomaly. Can anyone confirm this?

How can sending more data (>1MSS) help solving the problem?

Lastly, can anyone explain to me why this happens on IRIX (which is
"sort of" BSD-based), but not on the STREAMS-based Solaris? Does Solaris
has TCP_NODELAY set as default?

Thanks in advance,
Edwin.

Next message: Anne Wong: "HK FINE WINE SHOP ON NET"
Previous message: Rick Jones: "unsubscribing"
In reply to: Rick Jones: "Re: Different throughputs in two directions"
Next in thread: Rick Jones: "Re: Throughput deadlock"