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

[McIDAS #HPW-941884]: What am I doing wrong?

Hi Gilbert,

> All I can say is, wow.

All I can say is phew!

> Oh, and thanks for breaking my computers! ;-) ;-) ;-) ;-)

I try to be nice ;-)  Seriously though, that debug logging would break
data serving (and not just write a lot of log output to a file) was
and still is totally unexpected on my part.  I have not see this on
any other system.  Given that the McIDAS installations on the toplevel
ADDE servers like adde.ucar.edu and adde.cise-nsf.gov are 32-bit,
I do not think that this is simply a 32-bit issue.  It must be something
in a CentOS configuration that limits the size of "something".  I say
something since the size of the trce file was no where near the 2 GB
limit for 32-bit systems.

> Actually, when I reinstall everything on weather3, it will break again. I
> just did a backup last night to it, so it will miss that.

The fix is very quick:

- reinstall on weather3 from backu
- cd /home/mcidas/mcidas2009/src
- edit servutil.h and change the value from '1' to '0'
- run 'make' followed by 'make intstall.bin'

That will do it...

> Anyway, RHEL6
> got released Wednesday, CentOS 6, based on Fedora 12 or 13, will be out
> around the first of the year, I reckon. I will upgrade then and see if
> that still happens. But as long as weather and weather2 are fine, I'm
> golden.

Sounds good.

> AND...my CFDGONLY.BAT script is working again.


> Thanks again for all your
> help, Tom. Glad to see we both learned something here from this!

No worries, and I did learn something new.

> I wonder how many people are seeing this issue and not knowing it?

Probably not many if anyone.  Lots of users get their data off of one
of the servers that we run (adde.ucar.edu, adde.cise-nsf.gov, etc.).
The ones that run their own servers may not be serving the GINI imagery
from NOAAPort (via the NIMAGE datafeed), or they may be running on
machines where the problem does not show up.

> Oh, and yes, outside of my noaaport server, I'm still 32 bit here.

OK.  My view of the world nowadays is mostly 64-bit:  my workstation
at Unidata is 64-bit, and my laptop (Dell Studio 14 Intel I7 w/8 GB
RAM) is also 64-bit.  Off of the top of my head, I am not sure if the
VMWare virtual machine I am running on my laptop is 32-bit or 64-bit,
but I suspect that it is 32-bit.  The thing is that I do not try to
serve data from within this virtual machine to outside sources, so
any/all local datasets are accessed as LOCAL-DATA which, as you
know, worked AOK on weather2!?

The limitation must have something to do with a system limitation
processes exec'd by other processes (or run from xinetd).  The GINI ADDE
servers and subservers exec'd by 'mcserv' showed the problem while
those executed on behalf of the user for local datasets did not.
All very interesting...


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: HPW-941884
Department: Support McIDAS
Priority: Normal
Status: Closed

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.