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

20030915: pqutil race condition

This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.

  • Subject: 20030915: pqutil race condition
  • Date: Mon, 15 Sep 2003 09:13:00 -0600
Arthur,

> To: <address@hidden>
> From: "Arthur A. Person" <address@hidden>
> Subject: pqutil race condition
> Organization: Penn State
> Keywords: LDM-6 pqutil

The above message contained the following:

> I started up the following on a newley made queue before the ldm was
> started to monitor data collection and noticed that it was taking over 90%
> of the processor (while no data was coming in):
> 
>    pqutil -r -v -w /usr/local/ldm/pq/ldm.pq
> 
> The output from the command is:
> 
> Sep 15 13:21:31 pqutil: Feedtype set to ANY
> Sep 15 13:21:31 pqutil: Pattern set to ".*"
> Sep 15 13:21:31 pqutil: From time set to Thu Jan  1 00:00:00 1970
> Sep 15 13:21:31 pqutil: To time set to Tue Jan 19 03:14:07 2038
> Sep 15 13:21:31 pqutil: Cursor set to Mon Sep 15 13:21:31 2003
> Sep 15 13:21:31 pqutil: Cursor direction set to TV_GT
> Sep 15 13:21:31 pqutil: info.origin: ls1.meteo.psu.edu
> Sep 15 13:21:31 pqutil: info.arrival: Mon Sep 15 13:21:31 2003
> 
> Sep 15 13:21:31 pqutil: info.feedtype: EXP
> Sep 15 13:21:31 pqutil: info.seqno: 0
> Sep 15 13:21:31 pqutil: info.ident: TEST PRODUCT
> Sep 15 13:21:31 pqutil: From time set to Thu Jan  1 00:00:00 1970
> Sep 15 13:21:31 pqutil: To time set to Tue Jan 19 03:14:07 2038
> 
> Under the same conditions, "ldmadmin watch" takes almost zero processor.
> This is running on RH 9 with ldm 6.0.14.

Odd.  I don't see that behavior here.

What version of the LDM are you running?  What happens if you change the
command to
 
    pqutil -r -vl- -w /usr/local/ldm/pq/ldm.pq

> I also notice that when I start it with -r -v when data is coming in, that
> it seems to be taking as much processor as it can get also.
> 
> Is there some bug here using the -r or -v switch?  I used the -r switch
> since I assumed that would take fewer resources and was safer for the
> queue.

You used pqutil(1) correctly.

This will require further investigation.  I'll let you know.

> Arthur A. Person
> Research Assistant, System Administrator
> Penn State Department of Meteorology
> email:  address@hidden, phone:  814-863-1563

Regards,
Steve Emmerson