LDM Frequently Asked Questions
This page contains answers to some of the most frequently asked questions
about the LDM.
Table of Contents
What is the LDM?
The LDM is a distributed system for event-driven data distribution. It
consists of a suite of software to select, capture, process, and distribute
data products using a set of network client/server programs and their shared
protocols.
Who can use the LDM software?
The LDM software is freely available to anyone who wishes to use it. It is copyrighted
by the University
Corporation for Atmospheric Resesarch.
How do I get the LDM distribution?
The Unidata LDM software is now used by more than 150 universities/cooperating
agencies. The LDM software is currently freely available via anonymous
FTP in
source form or as
binaries for several supported platforms from the Unidata Program Center.
Should I get the source or binary distribution of the LDM?
Click here for help on deciding.
Why do I get the error "bin/ldmadmin: Command not found" when I run ldmadmin
If you install the binary release of the LDM, the path to the
perl binary (which interprets the ldmadmin
script) is based on the perl installation at the Unidata
Program Center. This might
not be correct for your site's setup. The first line of each perl script
in the LDM distribution starts with a line that looks something like:
#!/usr/local/perl
If your perl binary is not located in the path noted on that
line, modify the line to use the correct path.
(IRIX 6.1) Why does my server deny RPC access to others?
In IRIX, version 6.1, there is a file, /etc/config/portmap.options. The -a
option in this file disallows any host trying to do an RPC call to the
machine. You must explicitly allow any hosts that will connect to you. See
the man page for portmap for more information.
(IRIX) Why does my LDM die every Sunday at 4:30 am?
On IRIX systems, a disk defragger process runs every Sunday at
4:30 a.m. out of the root account. Because the product queue is a memory-mapped
I/O file, it cannot be defragmented on disk while it is mapped. If it is,
it corrupts the queue.
The solution is to either remove the root cron job that runs
/usr/etc/fsr, or to add a -m switch pointing to a file that does
not include the file system that contains the queue. This will allow other
file systems to be defragmented, ignoring the data file system. See the
man page for fsr for more information.
4/30/96: SGI now has a patch available that seems to fix the fsr
problem: patchSG0000870 - EFS filesystem roll-up patch for non-XFS systems.
Why do I get the error that another server is running on port 388 when
there is no other server active?
If the LDM server does not exit cleanly, port 388 may remain registered with
the portmapper, even if the server is no longer running. Use
rpcinfo to check this:
% rpcinfo -p
If port 388 is still registered, as root, use rpcinfo to deregister
the port:
% rpcinfo -d 300029 version
where version is the version number of the program that is running,
which will be 5 and/or 6.