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

Re: 20000831: Diagnosing data slowness at Penn State (new thread) (fwd)




===============================================================================
Robb Kambic                                Unidata Program Center
Software Engineer III                      Univ. Corp for Atmospheric Research
address@hidden             WWW: http://www.unidata.ucar.edu/
===============================================================================

---------- Forwarded message ----------
Date: Thu, 31 Aug 2000 17:11:10 -0400 (EDT)
From: Arthur A. Person <address@hidden>
To: Robb Kambic <address@hidden>
Subject: Re: 20000831: Diagnosing data slowness at Penn State (new thread)

Rob,

I just discovered that I've been pulling data from dusk.ssec.wisc.edu
instead of motherlode... I had switched last week to try and improve
delivery times while UCAR was having a networking problem and forgot to
switch back.  I don't know how much of an impact this will have, but FYI
I've switched back to motherlode now and will monitor. 

                                         Art.

On Thu, 31 Aug 2000, Robb Kambic wrote:

> On Thu, 31 Aug 2000, Unidata Support wrote:
> 
> > 
> > ------- Forwarded Message
> 
> > 
> > >From: "Arthur A. Person" <address@hidden>
> > >Subject: Diagnosing data slowness
> > >Organization: Penn State
> > >Keywords: 200008311839.e7VId2N08503 IDD motherlode network
> > 
> > Hi...
> > 
> > Our ldm for DDPLUS, HRS, etc from motherlode is running an hour behind
> > (i.e. at the limit) and I'm trying to determine why.  Whereas these are
> > slow, data from WSI and MCIDAS (for example) are on time.  Does this mean
> > the problem is a network bandwidth one to motherlode, or could it still be
> > a volume data issue from motherlode going through the ldm queues?  I see
> > there are an aweful lot of pqexpire entries in ldmd.log and I know a bunch
> > of NIDS data was added to the Noaaport feed in the last several days.
> > What's the best way to diagnose this problem?
> > 
> >                                           Thanks.
> > 
> >                                             Art.
> 
> Art,
> 
> It seems that the problem is local to psu, it's taking more time to go
> through the local psu network then to go across the country.  Looking at
> the motherlode logs, I see time elapsed to navier.
> 
> Aug 31 08:17:53 motherlode.ucar.edu navier(feed)[16361]: h_clnt_call:
> navier.meteo.psu.edu: 1: time elapsed
>   55.239236
> Aug 31 08:18:47 motherlode.ucar.edu navier(feed)[16361]: h_clnt_call:
> navier.meteo.psu.edu: BLKDATA: time e
> lapsed  32.236476
> 
> In my opinion the problem is still with the psu internal network.
> 
> Robb Jeff
> 
> 
> motherlode.ucar.edu% traceroute navier.meteo.psu.edu 
> traceroute to navier.meteo.psu.edu (128.118.28.2), 30 hops max, 40 byte
> packets
>  1  mlr-n13 (128.117.13.251)  0.615 ms  0.361 ms  0.519 ms
>  2  vbnsr-n2 (128.117.2.252)  0.544 ms  0.496 ms  0.524 ms
>  3  internetr-n243-104 (128.117.243.106)  0.614 ms  0.571 ms  0.758 ms
>  4  denv-abilene (128.117.243.126)  2.226 ms  1.995 ms  1.654 ms
>  5  kscy-dnvr.abilene.ucaid.edu (198.32.8.14)  12.083 ms  12.612 ms
> 12.669 ms
>  6  ipls-kscy.abilene.ucaid.edu (198.32.8.6)  21.521 ms  21.410 ms  21.815
> ms
>  7  clev-ipls.abilene.ucaid.edu (198.32.8.26)  27.962 ms  27.338 ms
> 27.400 ms
>  8  abilene-gp.psc.net (192.88.115.122)  31.625 ms  31.329 ms  32.020 ms
>  9  penn-state-vbns.psc.net (198.32.224.66)  41.545 ms  39.589 ms  41.578
> ms
> 10  Willard1-ATM9-0-0.1.gw.psu.edu (128.118.44.2)  41.675 ms  42.644 ms
> 38.831 ms
> 11  * * *
> 12  dk8000-1-rtvlan.ems.psu.edu (128.118.64.5)  45.553 ms  43.808 ms
> 69.343 ms
> 13  navier.meteo.psu.edu (128.118.28.2)  43.537 ms *  40.712 ms
> 
> 
> 
> 
> 
> 
> 
> 
> 
> 
> > Arthur A. Person
> > Research Assistant, System Administrator
> > Penn State Department of Meteorology
> > email:  address@hidden, phone:  814-863-1563
> > 
> > 
> > ------- End of Forwarded Message
> > 
> 
> ===============================================================================
> Robb Kambic                              Unidata Program Center
> Software Engineer III                    Univ. Corp for Atmospheric Research
> address@hidden                   WWW: http://www.unidata.ucar.edu/
> ===============================================================================
> 
> 
> 
> 

Arthur A. Person
Research Assistant, System Administrator
Penn State Department of Meteorology
email:  address@hidden, phone:  814-863-1563