20030929: 20030929: Lots of problems with my LDM connection

>From: "David Wojtowicz" <address@hidden>
>Organization: UIUC
>Keywords: 200309291535.h8TFZhk1020058 IDD

Hi Dave,

I am wondering if the problems that Christian is having is related to
the clock being so far off on flood.atmos.uiuc.edu.  To see the clock
offset and continuing drift on flood, see the folloing:

http://www.unidata.ucar.edu/cgi-bin/rtstats/iddstats_nc?IDS|DDPLUS+flood.atmos.uiuc.edu

I was meaning to contact you about this anyway along with the following:

- 'downpour' is still running ldm-6.0.10.  We would appreciate it if you
  upgraded this machine to the current LDM release, 6.0.14.  The main
  reason for this is that the connection strategy was changed between
  6.0.10 and 6.0.14.  6.0.14 clients figure out if the server from
  which they are requesting data is running an LDM-6 and then only settle
  for an LDM-6 connection if they are.  LDM-6.0.10 would try an LDM-6
  connection and then drop back to an LDM-5 connection if they didn't
  get connected in a short period of time.  A good example of when
  an LDM-5 connection would be established is when the sending LDM
  is stopped and restarted while downstreams are connected.

- 'squall' is sending data to more than one downstream site:

  aeolus.valpo.edu
  zelgadis.geol.iastate.edu
  pluto.met.fsu.edu

  'zelgadis' and 'pluto' relay the data received from squall to sites
  further downstream.

  pqbinstats stats show that 'squall' is running LDM-6.0.14, but we are
  receiving no real time stats from it.  The lack of realtime stats
  from an IDD relay node prevents us from tracing feed problems beyond
  the node that is not reporting data.

  Can you update the ~ldm/etc/ldmd.conf file to report real time stats
  to us?  Thanks!

> I see lots of log messages like these:
>
>Sep 29 04:37:18 flood io(feed)[15922]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 04:38:59 flood io(feed)[15926]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 04:40:27 flood io(feed)[15962]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 04:42:35 flood io(feed)[15967]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 04:46:43 flood io(feed)[15979]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 04:51:15 flood io(feed)[16025]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 04:52:43 flood io(feed)[16037]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 04:57:08 flood io(feed)[16049]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 05:00:18 flood io(feed)[16096]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 05:01:46 flood io(feed)[16111]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 05:07:22 flood io(feed)[16146]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 05:12:20 flood io(feed)[16187]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 05:12:21 flood io(feed)[16146]: up6.c:288: nullproc_6() failure to
>io.sca.uqam.ca: RPC: Unable to send; errno = Broken pipe
>Sep 29 05:14:25 flood io(feed)[16190]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 05:27:26 flood io(feed)[16273]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 05:49:58 flood io(feed)[16411]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 05:51:06 flood io(feed)[16411]: up6.c:288: nullproc_6() failure to
>io.sca.uqam.ca: RPC: Unable to send; errno = Broken pipe
>Sep 29 05:51:19 flood io(feed)[16449]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 05:56:33 flood io(feed)[16449]: up6.c:288: nullproc_6() failure to
>io.sca.uqam.ca: RPC: Unable to send; errno = Broken pipe
>Sep 29 05:56:35 flood io(feed)[16478]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 06:16:43 flood io(feed)[16629]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 06:28:59 flood io(feed)[16724]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 06:30:53 flood io(feed)[16767]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 06:43:19 flood io(feed)[16849]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 06:44:08 flood io(feed)[16849]: up6.c:288: nullproc_6() failure to
>io.sca.uqam.ca: RPC: Unable to receive; errno = Connection reset by peer
>Sep 29 06:44:38 flood io(feed)[16851]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 06:46:55 flood io(feed)[16860]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 06:50:01 flood io(feed)[16900]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 06:51:01 flood io(feed)[16900]: up6.c:288: nullproc_6() failure to
>io.sca.uqam.ca: RPC: Timed out
>Sep 29 06:52:26 flood io(feed)[16917]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 06:56:04 flood io(feed)[16926]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 06:58:05 flood io(feed)[16926]: up6.c:288: nullproc_6() failure to
>io.sca.uqam.ca: RPC: Unable to send; errno = Broken pipe
>Sep 29 06:58:07 flood io(feed)[16932]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 07:00:23 flood io(feed)[17005]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 07:02:34 flood io(feed)[17005]: up6.c:288: nullproc_6() failure to
>io.sca.uqam.ca: RPC: Unable to send; errno = Broken pipe
>Sep 29 07:02:43 flood io(feed)[17009]: topo:  io.sca.uqam.ca UNIDATA
>Sep 29 07:02:54 flood io(feed)[16932]: up6.c:288: nullproc_6() failure to
>io.sca.uqam.ca: RPC: Unable to send; errno = Broken pipe
>
>---------------------------------------------------------------
>David Wojtowicz, Sr. Research Programmer
>Dept of Atmospheric Sciences, University of Illinois
>address@hidden (217) 333-8390
>
>
>----- Original Message ----- 
>From: "Steve Emmerson" <address@hidden>
>To: <address@hidden>
>Cc: <address@hidden>; <address@hidden>
>Sent: Monday, September 29, 2003 9:15 AM
>Subject: 20030929: Lots of problems with my LDM connection
>
>
>> Hi Christian,
>>
>> > To: address@hidden
>> > From: =?ISO-8859-1?Q?Christian_Pag=E9?= <address@hidden>
>> > Subject: Lots of problems with my LDM connection
>> > Organization: UQAM
>> > Keywords: LDM connection
>>
>> The above message contained the following:
>>
>> > I have a problem with my LDM connection. I have a lot of ERROR:
>> > requester6.c:205: Connection to upstream LDM closed
>> > all over the place in my logs, so my latency is very high. It does that
>> > for all my ldm feeds, but traceroute shows 23 ms response from
>> > flood.atmos.uiuc.edu... What is error 205 in requester6.c ?
>>
>> "requester6.c:205" refers to file "server/requester6.c", line 205.
>>
>> The error message means that the downstream LDM process noticed that the
>> connection to the upstream LDM was closed for an unknown reason.
>>
>> The reason for the failure might be in the logfile of the upstream LDM
>> on host "flood.atmos.uiuc.edu".  The email address for the LDM contact
>> is <address@hidden>.  I suggest that you contact them and ask for the
>> corresponding logfile entries (I've copied this email to that address).
>>
>> > Christian Pagé
>> > address@hidden
>> > http://meteocentre.com/    http://meteoalerte.com/
>> >
>> > Etudiant au Doctorat en Sciences de l'environnement UQAM
>> > +1 514 987 3000 ext. 2376
>>
>> Regards,
>> Steve Emmerson
>>
>