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

20021112: dmraob.k and dmsyn.k hanging on weather2; LDM memory leak (cont.)



>From: Gilbert Sebenste <address@hidden>
>Organization: NIU
>Keywords: 200210050242.g952g0127088 McIDAS-XCD DMRAOB DMSYN

Gilbert,

re: additional memory leaks that can cause the XCD data monitors to grow
to the point that they can't malloc more memory - still looking for leaks

>OK. Thanks!

>Well, LDM is doing it too. Can you pass this on to all who are working on 
>it?

Can you tell me which LDM routine(s) are exhibiting this behavior?  If
it is rtstats, we already have a bug fix available, but it is in the
form of source code.  This means that you would have to build the LDM
from source using the bugfixed code to get past the rtstats memory leak
problem.  If you decide to do this, I suggest you download and build
LDM 5.2.1:

<login as 'ldm'>
ftp ftp.unidata.ucar.edu
  <user> anonymous
  <pass> your_full_email_address
  cd pub/ldm5
  binary
  get ldm-5.2.1.tar.Z
  get binstats.c
  get ldmsend.c
  quit

zcat ldm-5.2.1.tar.Z | tar xvf -
mv binstats.c ldmsend.c ldm-5.2.1/src/rtstats
cd ldm-5.2.1/src
./configure
make
make install
sudo make install_setuids

For reference, rtstats was added to the LDM distribution starting with 5.2.

By the way, I just did a long list of ~ldm on weather2, and I am seeing
an incredible number of hidden files that reference 'netcdf'.  Here is
an example:

cd ~ldm
ls -alt
total 4940
drwxr-xr-x   29 ldm      users      200704 Nov 12 13:28 ./
-rw-------    1 ldm      users           0 Nov 12 13:28 .tmp_netcdf.ZAKIOF
-rw-------    1 ldm      users           0 Nov 12 13:25 .tmp_netcdf.074jeH
-rw-------    1 ldm      users           0 Nov 12 13:24 .tmp_netcdf.ooRtMb
-rw-------    1 ldm      users           0 Nov 12 13:18 .tmp_netcdf.jCHHqT
-rw-------    1 ldm      users           0 Nov 12 13:18 .tmp_netcdf.lTX1HG
drwx------   17 ldm      users        4096 Nov 12 13:18 .mctmp/
-rw-------    1 ldm      users           0 Nov 12 13:04 .tmp_netcdf.Ybi29h
 ...

I don't know what these are being caused from, but, even though they
are all zero length, they will be a problem at some point.

Tom