[McIDAS #MYU-247334]: resurrection of ADDE server on papagayo

Hi Clint,

As promised, I am writing to report what I did on idd.unl.edu wrt
installing/configuring McIDAS.  Hopefully I will have the energy
to put down all that I did :-)

As 'root':

1) changed the setup for the user 'mcadde':

   - set it to not have login capability (it was /bin/bash)
   - set its HOME directory to be the same as 'mcidas',
     /home/mcidas (it was /home/mcadde)

2) added firewall rules (/etc/sysconfig/iptables) to allow access
   through ports 112 (McIDAS ADDE remote server) and 388 (ldm).

3) installed xinetd, configure it to start at (re)boot:

   yum install xinetd
   chkconfig --level 345 xinetd on
   /etc/init.d/xinetd start

4) install g77 using:

   yum install compat-gcc-34-g77 (or something like this; I can never remember)

   This was to install the shared library needed by the ldm-mcidas
   decoder nldn2md.  It was not needed for McIDAS as I built McIDAS using
   gfortran.

5) configure the McIDAS ADDE remote server for use:

   cd /home/mcidas
   sh ./mcinet2007.sh install mcadde

6) install the MySQL pieces needed for the McIDAS ADDE GRIB server:

   yum install mysql-server
   yum install mysql-client
   yum install mysql-devel

   The client and server pieces were already installed.  The
   development environment had not been installed and it was needed
   because McIDAS routines need it for ADDE GRIB serving.

7) change the password for the MySQL installation to 'xxxxx' as
   per instructions in http://lists.mysql.com/mysql/171366:

   service mysql stop
   mysqld_safe --skip-grant-tables &
   mysql -uroot mysql 
   mysql> use mysql;
   mysql> UPDATE user SET password=PASSWORD("xxxxx") WHERE user="root"; 
   mysql> FLUSH PRIVILEGES;
   -- exit the mysqld_safe invocation

   chkconfig --level 345 mysqld on
   /etc/init.d/mysqld start

As 'mcidas':

1) download 'mcunpack', 'mcinet2007.sh', and 'mcidasx2007.tar.gz' from
   the Unidata website into the /home/mcidas directory

2) unpack the distribution:

   export McINST_ROOT=/home/mcidas
   chmod +x mcunpack
   ./mcunpack

3) build McIDAS-X and McIDAS-XCD:

   -- edit .bash_profile and add:

# McINST_ROOT and VENDOR conforming shell defines
# needed to bulid McIDAS
McINST_ROOT=$HOME
VENDOR=-gfortran
export McINST_ROOT VENDOR

# NOTE: uncomment the next two lines that set MySQL_ROOT if you
#       want to build with MySQL support:
#       - leave the value blank if MySQL is installed in a standard
#         location
#       - specify the value as the root directory of the MySQL
#         installation if MySQL is not installed in a standard
#         location
MySQL_ROOT=TRUE
export MySQL_ROOT

if [ -e $HOME/admin/mcidas_env.sh ]; then
  . $HOME/admin/mcidas_env.sh
fi

unalias ls vi

   . .bash_profile

   cd mcidas2007/src
   make all
   make install.all

4) configure McIDAS using the v2007 'mcxconfig' script:

   cd $McINST_ROOT/workdata
   mcxconfig

   Answered all questions knowing which IDD datastreams were
   being ingested and which were to be decoded into McIDAS-usable
   formats.

5) change the read/write permissions on $McINST_ROOT/workdata so
   that 'ldm' can read write files there

6) create the MySQL database needed for McIDAS GRIB filing and
   ADDE GRIB file serving:

   cd $McINST_ROOT/workdata
   gribadmin makedb
   <pass> xxxxx


As user 'ldm':

1) incorporate ldmd.conf_mcidas actions created by 'mcxconfig'
   into ~ldm/etc/ldmd.conf.  This involves running 'xcd_run MONITOR'
   at LDM start and a single instance of pqact for McIDAS processing.

2) incorporated pqact.conf_mcidas[AB] actions produced by
   'mcxconfig' into a single pattern action file,
   pqact.conf_mcidas and put it in the ~ldm/etc/directory

3) incorporated crontab.ldm_mcidas entries for scouring of files
   produced by McIDAS-XCD and ldm-mcidas decoders (proftomd and
   nldn2md) into 'ldm's cron file

4) changed the read/write permissions on ~ldm/logs so that 'mcidas'
   can read/write it

5) as per instructions by 'mcxconfig', copied needed files from
   the McIDAS installation to the ~ldm/decoders and ~ldm/util
   directories editing them if/when needed to match the setup found.

6) stop and restart the LDM

7) install system performance monitoring via the Tcl script uptime.tcl
   which I installed in the ~ldm/util directory.  Add running of
   uptime.tcl into 'ldm's crontab.  The output file is ~ldm/logs/idd.uptime.
   You can look through this file to get a good idea of how hard the
   machine is working.

That is it as far as I can remember.

The things of note are:

- large number of NEXRCOMP (NEXRAD Level III composite images) and
  NIDS (NEXRAD Level III products for all NEXRADs)

- unusually high load averages consistently seen

Things to be done:

- contact your network people and ask them to configure DNS so that
  the machine can be know as both 'idd.unl.edu' and 'adde.unl.edu'.
  While at it, I would also request that the machine be knowable
  as 'thredds.unl.edu'.

I am keeping an eye on the machine to make sure that things are working
smoothly -- they are so far.

Please let me know if you (or your system admin) have any questions about
what I did or why I did them.

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: MYU-247334
Department: Support McIDAS
Priority: Low
Status: Closed