[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

20051101: IDD latencies (cont.)

>From: "Patrick L. Francis" <address@hidden>
>Organization: BGSU
>Keywords: 200510312116.j9VLGv7s008531 IDD

Patrick,

re: I recommend you contact your campus computing/networking group to
report the degraded service.

>already have..

Very good.

>all of my servers are wide open and not capped...
>
>traceroute from me to idd = good
>[ldm@weather]$ /usr/sbin/traceroute idd.unidata.ucar.edu
>traceroute to idd.unidata.ucar.edu (128.117.140.3), 30 hops max, 38 byte packe
> ts
> 1  router-112 (129.1.112.254)  0.375 ms  0.275 ms  0.240 ms
> 2  10.9.2.1 (10.9.2.1)  0.306 ms  0.340 ms  0.242 ms
> 3  10.9.10.12 (10.9.10.12)  0.377 ms  0.305 ms  0.268 ms
> 4  129.1.8.190 (129.1.8.190)  0.423 ms  0.418 ms  0.406 ms
> 5  gw-bgsu (129.1.8.253)  2.007 ms  2.791 ms  1.785 ms
> 6  toldb-r0-ge7-2s462.core.oar.net (199.18.159.17)  4.912 ms  5.221 ms  5.502 
> ms
> 7  clmbn-r1-po0-0.core.oar.net (199.18.145.17)  6.575 ms  7.098 ms  5.340 ms
> 8  clmbn-r0-po3-0.core.oar.net (199.18.145.1)  5.456 ms  5.861 ms  4.949 ms
> 9  clmbq-r2-po1-0.bb.oar.net (199.18.145.38)  5.277 ms  4.992 ms  5.376 ms
>10  abilene-IPLSng.ohio-gigapop.oar.net (192.88.192.134)  10.117 ms  9.849 ms  
>9.849 ms
>11  kscyng-iplsng.abilene.ucaid.edu (198.32.8.81)  28.877 ms  21.487 ms  19.03 
>4 ms
>12  dnvrng-kscyng.abilene.ucaid.edu (198.32.8.13)  29.700 ms  38.200 ms  29.82 
>0 ms
>13  192.43.217.129 (192.43.217.129)  30.180 ms  29.717 ms  30.094 ms
>14  gin.ucar.edu (128.116.254.240)  31.557 ms  31.274 ms  31.071 ms
>15  flrb.ucar.edu (128.117.243.83)  31.910 ms  30.957 ms  32.137 ms
>16  idd.unidata.ucar.edu (128.117.140.3)  30.817 ms  31.574 ms  30.828 ms
>[ldm@weather]$

'traceroute' listings are not a good measure of throughput you will get
using LDM/IDD/FTP/SCP/etc.  'traceroute' sends small packets and times
their round trip; realworld applications send large volumes of data
through the network.  Packet shaping software run by many university IT
departments is typically configured to limit the overall volume of a
stream, so if implemented so that it is controlling data flow on port
388 (LDM), you will see latencies climb.  A lot of times when there are
router problems, the effects are seen to be a direct function of
traffic volume.

>but both times after i mention it to suppor@unidata the peak drops without
>any changes on my end...

I imagine that the problem is not on your departmental machines, but,
rather, in the university network or in its connection to the Internet.

>http://www.unidata.ucar.edu/cgi-bin/rtstats/iddstats_nc?NIMAGE+unidata3.bgsu.e
> du

Yes, but compare this to other machines receiving the same data
from some of the same servers:

http://www.unidata.ucar.edu/cgi-bin/rtstats/rtstats_feedtree?NIMAGE

>perhaps the gremlins from Halloween are just a little overzealous this year.

:-)

Cheers,

Tom
--
NOTE: All email exchanges with Unidata User Support are recorded in the
Unidata inquiry tracking system and then made publicly available
through the web.  If you do not want to have your interactions made
available in this way, you must let us know in each email you send to us.