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

20040308: Logs indicate Latency problem: LSU to ULM

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: 20040308: Logs indicate Latency problem: LSU to ULM
  • Date: Tue, 09 Mar 2004 05:54:45 -0700
>From: Robert Leche <address@hidden>
>Organization: LSU
>Keywords: 200403081533.i28FXNrV006465 IDD latency

Bob and Adam,

Bob wrote:
>I wanted to let you know that I have taken notice of the RTSTATS Index 
>by Site for tornado.geos.ulm.edu. I see wide oscillations of 2 to 8 
>thousand second delays. This problem seems have reoccurred or started 
>last Thursday.  I was hoping the problem would 'go away' over the 
>weekend, but it has not.  I will initiate a trouble ticket and have our 
>LSU telecommunications group check into this.

I took a quick look at the latency plots for the IDS|DDPLUS feed
on tornado, and see the high latencies you are referring to:

http://my.unidata.ucar.edu/cgi-bin/rtstats/iddstats_nc?IDS|DDPLUS+tornado.geos.ulm.edu

I note, however, that tornado is redundantly feeding from seistan
and stokes.metr.ou.edu and the latencies from both feed paths are
equally bad.  To me this means that the problem is likely at ULM
and not necessarily at LSU.

Adam,

Can you check with your network folks to see if there is any known
problems at ULM?  Also, I would ask if packet shaping has been
implemented.  I suspect that there has been some kind of an
imposed slowdown since the latencies for feeds that have low number
of products (e.g., UNIWISC, FNEXRAD) are significantly better than for
feeds that have large numbers of products (e.g., IDS|DDPLUS, HDS).

Cheers,

Tom
--
NOTE: All email exchanges with Unidata User Support are recorded in the
Unidata inquiry tracking system and then made publically 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.

>From address@hidden  Fri Mar 12 13:00:49 2004

Last Thursday we starting taking in alot more nexrad data.  I will comment 
this out to see if it helps the latencys.  Also, we have completely maxed the 
I2 bandwidth out on campus and are in the process of aloting more (3 to 12).  
Another project that I am working on for the campus is a multicasting video 
conference room.  We starting testing on Thursday also.  That's where the 
maxing of the I2 came from.  With it off however we are below the maximum. 

Adam