[Met_help] [rt.rap.ucar.edu #90994] History for Issue with temporary met_config files

John Halley Gotway via RT met_help at ucar.edu
Fri Jul 19 13:52:08 MDT 2019


----------------------------------------------------------------
  Initial Request
----------------------------------------------------------------

Greetings Met Help
We are running into an issue where a user's home directory is filling up
with met_config_*****_0 files.  These files appear to be generated by
grid_stat.  Looking at the documentation and the code, this file should be
written to the tmp_dir defined in the grid_stat_config file.  That is
currently set to /tmp.  We have tried changing this as well as setting the
MET_TMP_DIR variable, but the met_config files continue to be written to
the home directory.  grid_stat is not being run from the user's home
directory.
If there is another setting that we are unaware of that will move the
location of these met_config files, please let us know.
Thanks
John


----------------------------------------------------------------
  Complete Ticket History
----------------------------------------------------------------

Subject: Issue with temporary met_config files
From: Julie Prestopnik
Time: Wed Jul 10 11:34:25 2019

Hi John.

I see that you are having difficulty with met_config files.  I have
assigned this ticket to John Halley Gotway who should be able to
better
assist you.

Julie

On Wed, Jul 10, 2019 at 5:56 AM John L Wagner - NOAA Federal via RT <
met_help at ucar.edu> wrote:

>
> Wed Jul 10 05:56:04 2019: Request 90994 was acted upon.
> Transaction: Ticket created by john.l.wagner at noaa.gov
>        Queue: met_help
>      Subject: Issue with temporary met_config files
>        Owner: Nobody
>   Requestors: john.l.wagner at noaa.gov
>       Status: new
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90994 >
>
>
> Greetings Met Help
> We are running into an issue where a user's home directory is
filling up
> with met_config_*****_0 files.  These files appear to be generated
by
> grid_stat.  Looking at the documentation and the code, this file
should be
> written to the tmp_dir defined in the grid_stat_config file.  That
is
> currently set to /tmp.  We have tried changing this as well as
setting the
> MET_TMP_DIR variable, but the met_config files continue to be
written to
> the home directory.  grid_stat is not being run from the user's home
> directory.
> If there is another setting that we are unaware of that will move
the
> location of these met_config files, please let us know.
> Thanks
> John
>
>

------------------------------------------------
Subject: Issue with temporary met_config files
From: John Halley Gotway
Time: Wed Jul 10 12:35:55 2019

Hi John,

I see you have a question about temp files created when running the
MET
tools.  Those "met_config_*" files are created when reading config
files in
MET.  The MET tools *should* be cleaning them up.  But I do see that
they
persist if there is a failure when parsing the config file.

I'm able to replicate this behavior, and I see that they are always
written
to the current working directory.  Setting the MET_TMP_DIR environment
variable has no effect, as you'd expect.  I ran via the debugger and
found
what's causing this behavior.

There are a couple lines of code that need to change to correctly
write
them to the location indicated by $MET_TMP_DIR environment variable...
or
"/tmp" by default if MET_TMP_DIR is not set.  I'll definitely include
this
change in the next release.  But if this is a problem, we could
include the
change in a bugfix version for met-8.1.

The good news is that once he gets the MET tools parsing the config
files
well, the tools should clean up those temp files on its own.

Please let me know if you need a bugfix or including the change in the
next
release is sufficient.

Thanks,
John





On Wed, Jul 10, 2019 at 11:38 AM Julie Prestopnik via RT
<met_help at ucar.edu>
wrote:

>
> Wed Jul 10 11:34:48 2019: Request 90994 was acted upon.
> Transaction: Given to johnhg (John Halley Gotway) by jpresto
>        Queue: met_help
>      Subject: Issue with temporary met_config files
>        Owner: johnhg
>   Requestors: john.l.wagner at noaa.gov
>       Status: open
>  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90994 >
>
>
> This transaction appears to have no content
>

------------------------------------------------
Subject: Issue with temporary met_config files
From: John L Wagner - NOAA Federal
Time: Wed Jul 10 13:42:42 2019

John
If you could send us a bugfix version to test, we would appreciate it.
Mike's testing has shown that the files will clean themselves up, but
he is
running enough instances that his home directory fills up with the
files
before they are able to be cleaned up.  We need to get these files
into
another directory and successfully setting $MET_TMP_DIR seems to be
the
only solution at the moment.
We'll continue to work on resolving any possible parsing issues that
may be
causing the met_config files to not be written to /tmp.
Thanks
John

On Wed, Jul 10, 2019 at 2:35 PM John Halley Gotway via RT
<met_help at ucar.edu>
wrote:

> Hi John,
>
> I see you have a question about temp files created when running the
MET
> tools.  Those "met_config_*" files are created when reading config
files in
> MET.  The MET tools *should* be cleaning them up.  But I do see that
they
> persist if there is a failure when parsing the config file.
>
> I'm able to replicate this behavior, and I see that they are always
written
> to the current working directory.  Setting the MET_TMP_DIR
environment
> variable has no effect, as you'd expect.  I ran via the debugger and
found
> what's causing this behavior.
>
> There are a couple lines of code that need to change to correctly
write
> them to the location indicated by $MET_TMP_DIR environment
variable... or
> "/tmp" by default if MET_TMP_DIR is not set.  I'll definitely
include this
> change in the next release.  But if this is a problem, we could
include the
> change in a bugfix version for met-8.1.
>
> The good news is that once he gets the MET tools parsing the config
files
> well, the tools should clean up those temp files on its own.
>
> Please let me know if you need a bugfix or including the change in
the next
> release is sufficient.
>
> Thanks,
> John
>
>
>
>
>
> On Wed, Jul 10, 2019 at 11:38 AM Julie Prestopnik via RT <
> met_help at ucar.edu>
> wrote:
>
> >
> > Wed Jul 10 11:34:48 2019: Request 90994 was acted upon.
> > Transaction: Given to johnhg (John Halley Gotway) by jpresto
> >        Queue: met_help
> >      Subject: Issue with temporary met_config files
> >        Owner: johnhg
> >   Requestors: john.l.wagner at noaa.gov
> >       Status: open
> >  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90994 >
> >
> >
> > This transaction appears to have no content
> >
>
>

------------------------------------------------
Subject: Issue with temporary met_config files
From: John Halley Gotway
Time: Wed Jul 10 14:22:38 2019

John,

OK, I committed this as a bugfix for met-8.1 to be included in our
next
patch release (met-8.1.2).  Here's the GitHub issue describing this
change:
https://github.com/NCAR/MET/issues/1161

You can find the single updated file here:
https://github.com/NCAR/MET/blob/master_v8.1/met/src/basic/vx_config/config_file.cc

You can use this version of the file to override
src/basic/vx_config/config_file.cc.  Do a make clean and recompile
before
testing the change.

If $MET_TMP_DIR is set, it'll write the met_config_* files to that
directory.  If not, it'll write them to /tmp instead.

Please let me know how it goes and confirm that the change works as
expected.

Thanks,
John

On Wed, Jul 10, 2019 at 1:42 PM John L Wagner - NOAA Federal via RT <
met_help at ucar.edu> wrote:

>
> <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90994 >
>
> John
> If you could send us a bugfix version to test, we would appreciate
it.
> Mike's testing has shown that the files will clean themselves up,
but he is
> running enough instances that his home directory fills up with the
files
> before they are able to be cleaned up.  We need to get these files
into
> another directory and successfully setting $MET_TMP_DIR seems to be
the
> only solution at the moment.
> We'll continue to work on resolving any possible parsing issues that
may be
> causing the met_config files to not be written to /tmp.
> Thanks
> John
>
> On Wed, Jul 10, 2019 at 2:35 PM John Halley Gotway via RT <
> met_help at ucar.edu>
> wrote:
>
> > Hi John,
> >
> > I see you have a question about temp files created when running
the MET
> > tools.  Those "met_config_*" files are created when reading config
files
> in
> > MET.  The MET tools *should* be cleaning them up.  But I do see
that they
> > persist if there is a failure when parsing the config file.
> >
> > I'm able to replicate this behavior, and I see that they are
always
> written
> > to the current working directory.  Setting the MET_TMP_DIR
environment
> > variable has no effect, as you'd expect.  I ran via the debugger
and
> found
> > what's causing this behavior.
> >
> > There are a couple lines of code that need to change to correctly
write
> > them to the location indicated by $MET_TMP_DIR environment
variable... or
> > "/tmp" by default if MET_TMP_DIR is not set.  I'll definitely
include
> this
> > change in the next release.  But if this is a problem, we could
include
> the
> > change in a bugfix version for met-8.1.
> >
> > The good news is that once he gets the MET tools parsing the
config files
> > well, the tools should clean up those temp files on its own.
> >
> > Please let me know if you need a bugfix or including the change in
the
> next
> > release is sufficient.
> >
> > Thanks,
> > John
> >
> >
> >
> >
> >
> > On Wed, Jul 10, 2019 at 11:38 AM Julie Prestopnik via RT <
> > met_help at ucar.edu>
> > wrote:
> >
> > >
> > > Wed Jul 10 11:34:48 2019: Request 90994 was acted upon.
> > > Transaction: Given to johnhg (John Halley Gotway) by jpresto
> > >        Queue: met_help
> > >      Subject: Issue with temporary met_config files
> > >        Owner: johnhg
> > >   Requestors: john.l.wagner at noaa.gov
> > >       Status: open
> > >  Ticket <URL:
https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90994
> >
> > >
> > >
> > > This transaction appears to have no content
> > >
> >
> >
>
>

------------------------------------------------
Subject: Issue with temporary met_config files
From: John L Wagner - NOAA Federal
Time: Fri Jul 19 13:02:29 2019

John
Sorry for the delay in reporting this, but I was able to confirm the
other
day that this did fix our problem.  I was able to set MET_TMP_DIR and
the
met_config_* files were indeed written then.
Thanks again for making the fix available.
John

On Wed, Jul 10, 2019 at 4:22 PM John Halley Gotway via RT
<met_help at ucar.edu>
wrote:

> John,
>
> OK, I committed this as a bugfix for met-8.1 to be included in our
next
> patch release (met-8.1.2).  Here's the GitHub issue describing this
change:
> https://github.com/NCAR/MET/issues/1161
>
> You can find the single updated file here:
>
>
https://github.com/NCAR/MET/blob/master_v8.1/met/src/basic/vx_config/config_file.cc
>
> You can use this version of the file to override
> src/basic/vx_config/config_file.cc.  Do a make clean and recompile
before
> testing the change.
>
> If $MET_TMP_DIR is set, it'll write the met_config_* files to that
> directory.  If not, it'll write them to /tmp instead.
>
> Please let me know how it goes and confirm that the change works as
> expected.
>
> Thanks,
> John
>
> On Wed, Jul 10, 2019 at 1:42 PM John L Wagner - NOAA Federal via RT
<
> met_help at ucar.edu> wrote:
>
> >
> > <URL: https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90994 >
> >
> > John
> > If you could send us a bugfix version to test, we would appreciate
it.
> > Mike's testing has shown that the files will clean themselves up,
but he
> is
> > running enough instances that his home directory fills up with the
files
> > before they are able to be cleaned up.  We need to get these files
into
> > another directory and successfully setting $MET_TMP_DIR seems to
be the
> > only solution at the moment.
> > We'll continue to work on resolving any possible parsing issues
that may
> be
> > causing the met_config files to not be written to /tmp.
> > Thanks
> > John
> >
> > On Wed, Jul 10, 2019 at 2:35 PM John Halley Gotway via RT <
> > met_help at ucar.edu>
> > wrote:
> >
> > > Hi John,
> > >
> > > I see you have a question about temp files created when running
the MET
> > > tools.  Those "met_config_*" files are created when reading
config
> files
> > in
> > > MET.  The MET tools *should* be cleaning them up.  But I do see
that
> they
> > > persist if there is a failure when parsing the config file.
> > >
> > > I'm able to replicate this behavior, and I see that they are
always
> > written
> > > to the current working directory.  Setting the MET_TMP_DIR
environment
> > > variable has no effect, as you'd expect.  I ran via the debugger
and
> > found
> > > what's causing this behavior.
> > >
> > > There are a couple lines of code that need to change to
correctly write
> > > them to the location indicated by $MET_TMP_DIR environment
variable...
> or
> > > "/tmp" by default if MET_TMP_DIR is not set.  I'll definitely
include
> > this
> > > change in the next release.  But if this is a problem, we could
include
> > the
> > > change in a bugfix version for met-8.1.
> > >
> > > The good news is that once he gets the MET tools parsing the
config
> files
> > > well, the tools should clean up those temp files on its own.
> > >
> > > Please let me know if you need a bugfix or including the change
in the
> > next
> > > release is sufficient.
> > >
> > > Thanks,
> > > John
> > >
> > >
> > >
> > >
> > >
> > > On Wed, Jul 10, 2019 at 11:38 AM Julie Prestopnik via RT <
> > > met_help at ucar.edu>
> > > wrote:
> > >
> > > >
> > > > Wed Jul 10 11:34:48 2019: Request 90994 was acted upon.
> > > > Transaction: Given to johnhg (John Halley Gotway) by jpresto
> > > >        Queue: met_help
> > > >      Subject: Issue with temporary met_config files
> > > >        Owner: johnhg
> > > >   Requestors: john.l.wagner at noaa.gov
> > > >       Status: open
> > > >  Ticket <URL:
> https://rt.rap.ucar.edu/rt/Ticket/Display.html?id=90994
> > >
> > > >
> > > >
> > > > This transaction appears to have no content
> > > >
> > >
> > >
> >
> >
>
>

------------------------------------------------


More information about the Met_help mailing list