Thank you, Steve, it was a pleasure to test this! I do think there is one other
bug for NOAAport downlink users that really needs to be mentioned:
Last November, a corrupted grib file crashed (technically, stopped) the LDM
nationwide at all NOAAport downlink sites at the NWS, .EDU's and .COM's (and
elsewhere). This can also happen if you have bad data reception of a grib file.
Now, it just throws an error message to the log, and it happily carries on.
It's just a flesh wound.
Additionally, a number of bugs which corrupted the queue have been swatted.
Most of them were rare, one of them was not, and I was bitten by that bug.
Finally, I must also acknowledge Mike Zuranski at the College of DuPage who
also helped me test this extensively on his servers. This way, we ran it on
both Debian and CentOS operating systems. We tested them on both satellite
ingesters, and relays.
If you run a NOAAport satellite ingester from your Novra or other receiver,
this is a ASAP must-install. If you have trouble with queue corruption, this is
an ASAP must-install, all in my opinion. Either way,
a ton of other small bugs have been swatted, and you can read those in the
README file, and you really should upgrade after testing it on your system.
Gilbert Sebenste
Consulting Meteorologist
AllisonHouse, LLC
> On Sep 30, 2020, at 1:20 PM, Steven Emmerson <emmerson@xxxxxxxx> wrote:
>
>
> Greetings LDM user,
>
> Version 6.13.12 of the LDM has been released and may be obtained through the
> LDM's home-page.
>
> The Release Notes detail the modifications. Perhaps the most important are
> the following:
> The use of the IS_ALIVE protocol message has been replaced with a fixed, 5
> minute timeout.
> The logging level for products that are not inserted into the product queue
> because they are either too old or were not requested has been changed from
> INFO to WARN in order to raise the visibility of this issue.
> A bug in the use of the pqact(1) state-file (which indicates the
> insertion-time of the last, successfully-processed product) has been fixed.
> pqact(1) PIPE and FILE actions that have not been used in 6 hours are now
> removed from the set of active actions and their file descriptors are closed.
> This has greatly reduced the time it takes to shut down the LDM system for
> those instances doing local processing with many such actions lacking the
> "-close" option.
> Thread-creation within noaaportIngester(1) has been modified to prevent the
> program from immediately terminating if the (optional) RHEL Insights Client
> package has been installed and prevents a thread from being created with a
> high priority. In this case the thread will execute with normal priority and
> the risk of missing NOAAPort packets will increase.
> If you have any issues with this or any other LDM version, please notify
> <support-ldm@xxxxxxxxxxxxxxxx>.
>
> I would like to take this opportunity to thank Gilbert Sebenste for his
> willingness to test this new version.
>
> Regards,
> Steve Emmerson
> LDM Developer
> _______________________________________________
> NOTE: All exchanges posted to Unidata maintained email lists are
> recorded in the Unidata inquiry tracking system and made publicly
> available through the web. Users who post to any of the lists we
> maintain are reminded to remove any personal information that they
> do not want to be made public.
>
>
> ldm-users mailing list
> ldm-users@xxxxxxxxxxxxxxxx
> For list information or to unsubscribe, visit:
> https://www.unidata.ucar.edu/mailing_lists/
