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

20030603: Upgraded LDM, lost MCIDAS

>From: Robert Leche <address@hidden>
>Organization: LSU
>Keywords: 200306031449.h53EnGLd009881 IDD Unidata-Wisconsin

Hi Bob,

>When Sirocco's LDM was upgraded, we lost the MCIDAS feeds. While Sirocco 
>does not decode these feeds, the downstream feeders decode it and they 
>are not receiving MCIDAS data (Seistan,and via Seistan; Hurricane). No 
>Satelite images.

I just looked at volume plots of UNIWISC (aka MCIDAS in LDM-5.x)
imagery data on the inject machine (unidata2.ssec.wisc.edu), the UCAR
top level IDD relay machine (thelma.ucar.edu), and on sirocco
(sirocco.srcc.lsu.edu).  The plots which are all available from links
in the the real time statistics page:


show that the volume of UNIWISC data received by thelma is the same as
is injected at unidata2.ssec.wisc.edu, AND the volume of UNIWISC data
received at sirocco.srcc.lsu.edu is the same as that received by
thelma.  Since the real time statistics are generated on the local
machine (e.g., sirocco), it has to mean that you are, in fact,
receiving the image data.  The question now is why it is not being

>We do receive the radar images, models, and ddplus 
>data, but no Satelite images. The down stream feeders are still using 
>LDM 5.2.2. What is strange is everything but the satelite images works.

What do the log files on sirocco and its downstream feed sites say with
regards to the relay of UNIWISC data?

>What has changed?

A lot has changed in the LDM, but the ability to relay UNIWISC data is
not one of them.


What are the operating systems being run on the machines that are not
able to feed MCIDAS image data from sirocco?

>1) The Operating system on Sirocco was upgraded to Redhat 7.3 from 7.1. 
>The LDM version 6.011 from version 5.22.

This should be fine.  There are numerous sites running the LDM on
RedHat 7.3 Linux systems, and they are apparently not having a problem
receiving the UNIWISC data.

>Sirocco's ldmd.conf and pqact.conf file are attached. The ldmd and pqact 
>configure file incorparte the changes in the new version of LDM (I think!).
Did you happen to change the ldmd.conf file on the down stream machines
that are no longer getting the UNIWISC data?  What I have in mind is a
change in their ldmd.conf file that renamed MCIDAS to UNIWISC.  This
would be incorrect in LDM-5.x.  It takes an LDM-6 to recognize UNIWISC
as a valid feed name.

>Is Thelma still my correct general IDD feed source?

If sirocco is going to be the top level IDD relay node, then its
primary feed for NOAAPORT data will be the SDI ingester that you
operate, datoo.srcc.lsu.edu.  We have been waiting for you to complete
upgrades of LSU machines to a current release of LDM-6 before advising
you of how we would like it configured for redunancy of IDD feeds.

Can you give us an estimate of when you will do the upgrade of your
other machines?

>Is Thelma still my correct feed source of MCIDAS data?

If sirocco is going to be the top level IDD relay node at LSU, then it
would be better to feed UNIWISC from its source, unidata2.ssec.wisc.edu.

This will not, however, change the receipt of UNIWISC data on sirocco
since we believe that you are, in fact, getting it already.

>Related issue:
>The LDM logs on Sirocco are not correct since upgrading. None of the 
>yyymmddhh.stats in logs/ are updating. Did I configure the syslog file 
>correctly? It is attached.

Here is the LDM entry from your /etc/syslog.conf file:

local0.debug       /usr/local/ldm/logs/ldmd.log

Assuming that your LDM log file is located in
/usr/local/ldm/logs/ldmd.log, and assuming that the white spaces
between 'local0.debug' and /usr/local/ldm/logs/ldmd.log are spaces and
not tabs, then I think you need to edit /etc/syslog.conf and change the
white space from spaces to tabs.  Also, it is unclear if lines above
this entry in /etc/syslog.conf are blank (no characters on the line
_including_ no spaces).  If there are empty lines above this entry that
_do_ have spaces in them, then you should either eliminate the spaces,
or delete the lines.

After you make changes to /etc/syslog.conf, you have to send a HUP
signal to syslogd to tell it to reread the configuration file.  In some
extreme cases (old Solaris systems where there were syslogd bugs),
syslogd sometimes has to be killed and restarted.  Again, this is
only rarely the case.

>Unrelated issue:
>Do you guys have the LDM administrator contact information for 

Like Jeff sent in a separate email, the last contact we had for Jackson
State was:

Duanjun Lu <address@hidden>

I don't think this is valid since:

% telent twister.jsums.edu 25
Connected to twister.jsums.edu.
Escape character is '^]'.
220 twister.jsums.edu ESMTP Sendmail 8.12.8/8.12.5; Tue, 3 Jun 2003 12:53:24 

vrfy dlu


252 2.5.2 Cannot VRFY user; try RCPT to attempt delivery (or try finger)