Mike wrote:
> Dividing the tcpstat.d output values by 10 brings them in line with the
rest of your measurements. Are you sure you didn't make a 10x error when
comparing your modified tcpstat.d with the original? (hint: the key probe
is the last one in the script where the counters are reset).
Uhmm.
Yes. You´re right, kind of :)
Thank you!
I tried again with the regular script:
TCP_out TCP_outRe TCP_in TCP_inDup TCP_inUn
100 0
107159168 0 0
316 0
103296672 0 0
100 0
105086552 0 0
612 0
102532224 0 0
100 0
102800916 0 110048
100 0
107525452 0 0
100 0
101727804 0 111000
100 0
108321680 0 0
100 0
104378992 0 116792
100 0
104001056 0 111000
100 0
105311944 0 122408
100 0
106907924 0 0
100 0
105918264 0 0
612 0
98864416 0 0
100 0
104936868 0 0
100 0
108334572 0 0
100 0
106463980 0 0
100 0
109640400 0 0
100 0
103656656 0 124032
100 0
104951880 0 121136
Now we have 105311944 = 105,311,944 which is 100 million.
But this is still too much compared to the the other tools, which show
about 70 to 80 MB/s, which seems to right for a 1Gb link.
Am i doing something wrong?
Or do i missinterprete the output of the script?
Thanx for any further enlightenment :)
[ Tolle Angebote findet ihr unter: http://www.uni.de/shopping ]
_______________________________________________
dtrace-discuss mailing list
dtrace-discuss@opensolaris.org
