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

[LDM #NQT-975754]: Problems with LDM

Hi Rob,

I apologize for the tardy reply...

re:
> I have been having problems establishing a data feed using LDM. I am
> running LDM 6.6.5 on a Linux box within our labs firewall. However, the
> firewall on my box is disabled. In any event, I'm not sure if the
> firewall is interferring or what.

The following line in your ldmd.log file indicates that your firewall
should not be a problem:

Aug 12 00:30:21 nexrad castor.itap.purdue.edu[6451] NOTE: Upstream LDM-6
on castor.itap.purdue.edu is willing to be a primary feeder

> I have tried many things, including
> turning on portmap, adding entires to hosts.allow, reinstalling LDM (to
> be sure I didn't screw up the first time).

Did you make sure to do the last LDM installation step as 'root':

-- download LDM to ~ldm
zcat ldm-6.6.5.tar.gz | tar xvf -
cd ldm-6.6.5/src
./configure
make
make install
sudo make install_setuids      <- 'make install_setuids' must be done as
                                  'root' or with 'root' privilege

cd
rm -f runtime
ln ldm-6.6.5 runtime
ln -s runtime/* .

To check, please send the output of:

cd ~ldm
ls -alt bin/*

> I have the following line in my ldmd.conf file:
> 
> REQUEST NEXRAD2 ".*" castor.itap.purdue.edu

This looks fine.  The ldmd.log message that shows that castor.itap.purdue.edu
is willing to be a primary feeder shows that your LDM successfully contacted
castor.

> 
> I am consistently getting the following error in my log file:  "RPC:
> Remote system error - No route to host." The log file is shown below

This is bothersome since it indicates some sort of network error.
What is the output of the following when you see the 'No route to host'
message:

ldmping castor.itap.purdue.edu

traceroute castor.itap.purdue.edu

re:
> [ldm@nexrad logs]$ more ldmd.log
> Aug 12 00:30:18 nexrad rpc.ldmd[6443] NOTE: Starting Up (version: 6.6.5;
> built: Aug 10 2007 17:32:29)
> Aug 12 00:30:18 nexrad rpc.ldmd[6443] NOTE: Using local address
> 0.0.0.0:388
> Aug 12 00:30:21 nexrad rpc.ldmd[6443] NOTE: local_portmapper_running():
> clnttcp_create() failure: : RPC: Remote system error - No route to host
> Aug 12 00:30:21 nexrad pqact[6449] NOTE: Starting Up
> Aug 12 00:30:21 nexrad castor.itap.purdue.edu[6451] NOTE: Starting
> Up(6.6.5): castor.itap.purdue.edu:388 20070811233021.711 TS_ENDT
> {{NEXRAD2,  ".*"}}
> Aug 12 00:30:21 nexrad pqact[6449] WARN: No such file or directory
> Aug 12 00:30:21 nexrad castor.itap.purdue.edu[6451] NOTE: Previous
> product-information file ".77a625209c984eb24b6d25f78f11c0a7.info"
> doesn't exist
> Aug 12 00:30:21 nexrad pqact[6449] WARN: stateRead(): Couldn't open
> "/usr/local/ldm/etc/pqact.conf.state"
> Aug 12 00:30:21 nexrad pqact[6449] WARN: Previous-state information
> doesn't exist
> Aug 12 00:30:21 nexrad castor.itap.purdue.edu[6451] NOTE: LDM-6 desired
> product-class: 20070811233021.711 TS_ENDT {{NEXRAD2,  ".*"}}
> Aug 12 00:30:21 nexrad rtstats[6450] NOTE: Starting Up (6443)
> Aug 12 00:30:21 nexrad castor.itap.purdue.edu[6451] NOTE: Upstream LDM-6
> on castor.itap.purdue.edu is willing to be a primary feeder

It is curious that contact was established with the upstream feed host
in spite of the 'No route to host' message.  The output of traceroute
may shed some light on this.

> The netstat command issues the following information
> 
> 
> [root@nexrad ~]# netstat -a inet -p tcp | awk 'NR<=2 || /ldm/'
> Active Internet connections (servers and established)
> Proto Recv-Q Send-Q Local Address               Foreign Address
> State       PID/Program name
> tcp        0      0 *:ldm                       *:*
> LISTEN      6821/rpc.ldmd
> tcp        0     28 nexrad:32789
> castor.itap.purdue.edu:ldm  ESTABLISHED 6824/rpc.ldmd
> unix  2      [ ]         DGRAM                    18588  6821/rpc.ldmd

Like the message in ldmd.log, this indicates that your LDM is connected
to castor as a downstream.

> (where ldm=388)
> Looks like a connection is established between my machine (nexrad) and
> castor.itap.purdue.edu through port 32789. However, ldm is listening on
> port 388(=ldm). If understand this right, ldm is not going to "hear"
> anything.

No, this output is what is to be expected.  The port used on the
local machine is random; the port in use on the upstream machine
is 388.

> I can change the port that ldm listens to by modifying the
> ldmadmin-pl.conf file. But the port assignment for the connection
> changes everytime you restart ldm. So you never know what it will be.

You do not need to change the port on your system.

Is your clock properly set?  I ask since if your clock time is in the
future, your request for data from castor would be fore data that it
does not have.  In this case, the amount of time that your clock was
running in the future would have to elapse before castor would send
you data.

> Help!

Please let us know the results of ldmping and traceroute above.  Also,
please verify that your system clock is properly set and maintained
by ntpd.

Again, I apologize for the tardy reply.  Our LDM developer was on travel
since you first submitted your inquiry, and I was consumed by hosting
a McIDAS training workshop at the same time.

Cheers,

Tom
****************************************************************************
Unidata User Support                                    UCAR Unidata Program
(303) 497-8642                                                 P.O. Box 3000
address@hidden                                   Boulder, CO 80307
----------------------------------------------------------------------------
Unidata HomePage                       http://www.unidata.ucar.edu
****************************************************************************


Ticket Details
===================
Ticket ID: NQT-975754
Department: Support LDM
Priority: Urgent
Status: Closed