20051212: LDM - failure building 10G product queue

>From: Paul Prestopnik <address@hidden>
>Organization: NCAR/RAL
>Keywords: 200512122354.jBCNsl7s024279 LDM large queue

Hi Paul,

Missing from the information you included was what kind of hardware you
are using and what Operating System you are running (I assume Linux, but
we need more information).  Please send the output of the following:

uname -a

cat /proc/cpuinfo

LDM queues larger than 4 GB can be build on Linux systems that are
running on 64-bit hardware (e.g., AMD Opteron and Athlon 64 or Intel
Xeon EM64T) and are running the 64-bit version of the operating
system.  As soon as we determine that your system meets these criteria,
we can help you create a large LDM queue.

Cheers,

Tom Yoksas

>When we run the ldmadmin pqcreate -f option we get this error:
>
>> ldmadmin mkqueue -f
>Dec 12 23:48:37 pqcreate[2474] ERROR: ctl_init(): Data portion of file is too 
> big for one region
>pqcreate: create "/home/ldm/data/ldm.pq" failed: File too large
>Dec 12 23:48:37 UTC awl.rap.ucar.edu : make_pq: mkqueue failed
>
>Here are the values in our ldmadmin-pl.conf file:
>> ldmadmin config
>
>hostname:      awl.rap.ucar.edu
>os:            Linux
>release:       2.6.12-ral-smp
>ldmhome:       /home/ldm
>bin path:      /home/ldm/bin
>conf file:     /home/ldm/etc/ldmd.conf
>log file:      /home/ldm/logs/ldmd.log
>numlogs:       7
>log_rotate:    1
>data path:     /home/ldm/data
>product queue: /home/ldm/data/ldm.pq
>queue size:    10G bytes
>queue slots:   default
>IP address:    all
>port:          388
>PID file:      /home/ldm/ldmd.pid
>LDMHOSTNAME:   awl.rap.ucar.edu
>PATH:          /home/ldm/bin:/bin:/usr/bin:/usr/sbin:/sbin:/usr/ucb:/usr/usb:/
> usr/etc:/etc:.:/home/ldm/bin:/home/ldm/util:/home/ldm/decoders:/home/ldm/rap/
> bin:/home/ldm/decoders:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbi
> n:/bin:/usr/bin/X11
>
>
>Top shows that we have 12G of ram:
>
>top - 16:51:04 up 10 days, 38 min,  2 users,  load average: 0.06, 1.26, 1.14
>Tasks:  83 total,   1 running,  82 sleeping,   0 stopped,   0 zombie
>Cpu(s):  0.0% us,  0.0% sy,  0.0% ni, 99.5% id,  0.5% wa,  0.0% hi,  0.0% si
>Mem:  12474024k total,  4345580k used,  8128444k free,   182620k buffers
>Swap:  4200988k total,        0k used,  4200988k free,   112500k cached
>
>
>And df shows that we have 17G of space on /d1
>> df -h
>Filesystem            Size  Used Avail Use% Mounted on
>/dev/sda1             4.0G  1.5G  2.3G  39% /
>tmpfs                 6.0G  4.0K  6.0G   1% /dev/shm
>/dev/sda5             4.0G  142M  3.7G   4% /var
>/dev/sda6             4.0G   44M  3.8G   2% /home
>/dev/sda7              18G   37M   17G   1% /d1
>linux:/exports/linux-i386
>                      118G   43G   70G  38% /var/autofs/mnt/linux-i386
>
>note that ~ldm/data points to /d1/ldm/pq/
>
>awl:~> ll ~ldm/data
>lrwxrwxrwx  1 ldm ldm 10 Dec  5 14:36 /home/ldm/data -> /d1/ldm/pq/
>
>
>We would like to build a 10G queue, but we can't get one to succesfully build
>for sizes greater than 4G.
>
>-Paul Prestopnik
>
>
--
NOTE: All email exchanges with Unidata User Support are recorded in the
Unidata inquiry tracking system and then made publicly available
through the web.  If you do not want to have your interactions made
available in this way, you must let us know in each email you send to us.