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

20050228: IDD Latency issues (cont.)

>From: Jeff Weber <address@hidden>
>Organization: UCAR/Unidata
>Keywords: IDD latency NOAAPORT 

Hi Mark,

I am curious about your comment that jackie.unidata.ucar.edu is seeing
latencies.  jackie is an inject-only node on the IDD.  The latencies it
sees are as close to 0.00 as possible.  Were you looking at a different
machine and thinking it was jackie?  If the real time stats pages were
showing high latencies for jackie at any point, it would most likely
indicate come kind of a problem with the real time stats reporting,
not with ingest on jackie.

To expand on Jeff's comment regarding jackie, we are running two DVB-S
NOAAPORT boxes here at UCAR: jackie (dual 500 Mhz PIII running Fedora
Core 1 Linux) and noaaport (dual 550 Mhz PIII running Solaris x86
5.9).  We are also running a DVB-S ingester at LSU, samoon (single 450
Mhz PIII running Fedora Core 3 Linux), and SSEC is running two boxes in
Madison, dvb1 and dvb2 (both dual 3.2 Ghz Xeon running RH Enterprise WS
3).  jackie and desi have been essentially powering the NOAAPORT
component of the IDD for about two weeks (with backup from
atm.geo.nsf.gov).

Cheers,

Tom

re: Jeff Weber wrote:
>Yup, jackie is our test DVB-S ingest, we are bringing it up and down for
>testing..not a valid comparison. thelma is a good comp.
>
>stokes looks much better know..there were latencies in the 1000's of
>seconds on Friday, now just a few 100's at the beginning of the chart..
>
>
>Not sure what was going on, but it appears to have cleared up..
>
>Thanks for checking into it!
>
>Jeff
>---------------------------------------------------------------------
>Jeff Weber                                    address@hidden        :
>Unidata Program Center                        PH:303-497-8676        :
>University Corp for Atmospheric Research      3300 Mitchell Ln       :
>http://www.unidata.ucar.edu/staff/jweber      Boulder,Co 80307-3000  :
>---------------------------------------------------------------------
>
>On Fri, 25 Feb 2005, Mark J. Laufersweiler wrote:
>
>> None on this end. I restarted hte ldm on stokes with a new queue
>> and was waiting to hear what that may have done.
>>
>> As far as I know there have not been any changes, but that does
>> not mean anything around here since I am usually the last to hear
>> of any troubles.
>>
>> But from what I am seeing on the website, jackie.unidata.ucar is
>> seeing latencies as well.
>>
>> mjl
>>
>>
>> On Fri, 25 Feb 2005, Jeff Weber wrote:
>>
>> > Hi Mark,
>> >
>> > U of Northern Iowa is experiencing latencies, and they appear to be
>> > originating at stokes....?
>> >
>> > IDS|DDPLUS
>> >
>> > http://my.unidata.ucar.edu/cgi-bin/rtstats/iddstats_nc?IDS|DDPLUS+stokes.m
> etr.ou.edu
>> >
>> > and
>> >
>> > NNEXRAD
>> >
>> > http://my.unidata.ucar.edu/cgi-bin/rtstats/iddstats_nc?NNEXRAD+stokes.metr
> .ou.edu
>> >
>> >
>> > via our machines atm and thelma
>> >
>> > but atm does not show latency:
>> >
>> > IDS|DDPLUS
>> >
>> > http://my.unidata.ucar.edu/cgi-bin/rtstats/iddstats_nc?IDS|DDPLUS+atm.geo.
> nsf.gov
>> >
>> > NNEXRAD
>> >
>> > http://my.unidata.ucar.edu/cgi-bin/rtstats/iddstats_nc?NNEXRAD+atm.geo.nsf
> .gov
>> >
>> >
>> > Are you aware of any network issues since the 23rd?
>> >
>> > Cheers,
>> >
>> >
>> > Jeff
>> > ---------------------------------------------------------------------
>> > Jeff Weber                                    address@hidden        :
>> > Unidata Program Center                        PH:303-497-8676        :
>> > University Corp for Atmospheric Research      3300 Mitchell Ln       :
>> > http://www.unidata.ucar.edu/staff/jweber      Boulder,Co 80307-3000  :
>> > ---------------------------------------------------------------------
>> >
>>
>> --
>> --------------------------------------------------------------------
>>   Dr. Mark J. Laufersweiler     |
>>   School of Meteorology         |     Those who give up essential
>>   University of Oklahoma        |   liberties for temporary safety
>>   address@hidden                | deserve neither liberty nor safety.
>>   (405) 325-6032                |             - Benjamin Franklin
>>   (405) 325-7689  (fax)         |
>>
>
--
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.

>From address@hidden  Mon Feb 28 10:19:02 2005

I may be a bit color blind and read the graphic wrong from Friday.
jackie is not even appearing on the graph today and it may have
just been a restart at NIU.

OU for the past week (now almost 2 weeks) have been under attach
from outside, with many MS machines (and a few solaris) being
compromised. While not much network traffic was coming into OU, the
outbound was huge. Apparently OU was part of a few DDOS attacks and
mass spammings. That may have be a cause for the latencies, but I
would have expected other of the SoM downstreams to have shown those
latencies as well.

As I told Jeff, I am the last to be told if there has been any major
changes to the OU network configurations.

mjl