[c-nsp] RTR/SAA on smaller platforms

[Shakeel Ahmad]

a simple solution (which you may have to pay for) is to use Cisco IPM
(Internetwork Performance Monitor) ; which manages whole issues for you -
all configuration mess is managed by the IPM server (WIN/*NIX) and get a
best (which i have yet seen) GUI display of statistics for management ;-)





On 8/12/06, Clinton Work <clinton at scripty.com> wrote:
>
>
> Config is for a pair of 3550s running 12.1EA. The SAA probe will send 10
> UDP
> probes per second using DSCP 40 for the entire hour. UDP probe interval is
> 100ms and UDP payload is 160 bytes. The num-packets 590 tells SAA to send
> 590 probes every 60 seconds spaced 100ms apart.
>
> ! Router B - Loopback0 is 192.168.2.100
> conf t
> rtr responder
> end
>
> ! Router A - Loopback0 is 192.168.1.23
> conf t
> rtr 1
> type jitter dest-ipaddr 192.168.2.100 dest-port 16252 source-ipaddr
> 192.168.1.23 interval 100 num-packets 590
> tos 0xA0
> threshold 100
> timeout 500
> request-data-size 160
> exit
> rtr schedule 1 life 200000000 start-time now
> end
>
>
> Grab "show rtr collection" stats once an hour via the command line or via
> SNMP to check the results.
>
> Command line:
>
> successful probes = NumOfRTT
> average latency = RTTSum / NumOfRTT
>
> failed probes = PacketLossSD + PacketLossDS + PacketOutOfSequence +
> PacketMIA + PacketLateArrival
>
> max jitter = max of ( MaxOfPositivesSD, MaxOfNegativesSD,
> MaxOfPositivesDS,
> MaxOfNegativesDS  )
>
> The jitter measurement is a rough estimate and is not done according to
> RFC
> 1889 section A.8.
>
>
> routera# show rtr collection 1
>         Collected Statistics
>
> Entry Number: 1
> Target Address: 192.168.2.100, Port Number: 16252
> Start Time: 00:42:47.000 pdt Thu Aug 10 2006
> RTT Values:
> NumOfRTT: 35397 RTTSum: 125736  RTTSum2: 460744
> Packet Loss Values:
> PacketLossSD: 0 PacketLossDS: 0
> PacketOutOfSequence: 0  PacketMIA: 3    PacketLateArrival: 0
> InternalError: 0        Busies: 0
> Jitter Values:
> NumOfJitterSamples: 35337
> MinOfPositivesSD: 1     MaxOfPositivesSD: 13
> NumOfPositivesSD: 3047  SumOfPositivesSD: 3117  Sum2PositivesSD: 3685
> MinOfNegativesSD: 1     MaxOfNegativesSD: 13
> NumOfNegativesSD: 3011  SumOfNegativesSD: 3073  Sum2NegativesSD: 3635
> MinOfPositivesDS: 1     MaxOfPositivesDS: 13
> NumOfPositivesDS: 4075  SumOfPositivesDS: 4181  Sum2PositivesDS: 5139
> MinOfNegativesDS: 1     MaxOfNegativesDS: 14
> NumOfNegativesDS: 4125  SumOfNegativesDS: 4237  Sum2NegativesDS: 5261
> Interarrival jitterout: 0       Interarrival jitterin: 0        One Way
> Values:
> NumOfOW: 4916
> OWMinSD: 1      OWMaxSD: 4      OWSumSD: 10528  OWSum2SD: 25920
> OWMinDS: 1      OWMaxDS: 7      OWSumDS: 7578   OWSum2DS: 14166
>
>
>
> Jee Kay wrote:
> > On 11/08/06, Clinton Work <clinton at scripty.com> wrote:
> >> I have done a lot of SAA icmpecho and jitter probes on both the Cisco
> >> 1721
> >> and Catalyst 3550 switches.  They work quite well and I haven't had any
> >> issues. If you summarize the jitter statistics they can provide round
> >> trip
> >> latency,max jitter, and packet loss with one test. The jitter probe
> does
> >> require another Cisco device at the other end with "rtr responder"
> >> enabled.
> >
> > Could you possibly provide some sample configurations for how you were
> > doing this?
> >
> > How were you collecting the results?
> >
> > Thanks,
> > Ras
>
> --
> ===================================================
> Clinton Work    clinton at scripty.com
> Calgary, AB
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/
>