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

[LDM #JFS-571343]: LDM stops writing data to disk

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 Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated.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: [LDM #JFS-571343]: LDM stops writing data to disk
  • Date: Mon, 01 Apr 2019 16:55:27 -0600
Jim,

> I have been running the 6.13.6 version of LDM for 9 months without a
> single problem. Last week, without any user intervention or changes in
> the box stopped writing data to the disk. I can use ldmadmin watch and
> see data flowing and I can see the var/queue/ldm.pq file is updating time
> and size but it does not write any data out. Once it starts this behavior
> ldmadmin stop will not shutdown the ldm. If I reboot and clean the queue
> it will not work. If I delqueue and mkqueue it will come back up and run
> for between 10 to 50 minutes before stopping the writes to disk.  I am
> getting all of my data via NOAAPORT and I have read the support emails
> for days seeing many similar problems as mine, yet nothing mentioned in
> those corrects my issue. I tried moving to 6.13.10 but I get the same
> issue. I am not seeing any reason in any logs to what is going on. It
> is as if pqact is running, it just decides not to process anything from
> the queue anymore.
> 
> Is there any suggestions you can offer me of where I can look to solve
> this?

I've had one other LDM user report a similar problem. Unfortunately, we haven't 
seen that problem here.

It's as if a pqact(1) process hangs. It would be very helpful, therefore, if I 
could learn where the hung pqact(1) process is when this happens. Can you 
rebuild and re-install your LDM will debugging enabled by specifying the option 
"--enable-debug" to the configure(1) script. Then, when this happens, attach 
the debugger to the hung pqact(1) process via the "-p <pid>" option of gdb(1) 
and get a stack trace via the command "where". You can then quit the debugging 
session via the command "q".

I would love to see that stack trace.

Regards,
Steve Emmerson

Ticket Details
===================
Ticket ID: JFS-571343
Department: Support LDM
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.