[ncl-talk] NCL addfile memory leak in functions
Ghassan Alaka - NOAA Federal
ghassan.alaka at noaa.gov
Fri Dec 13 10:01:20 MST 2019
Well thank you for the sanity checks. Rick: I think there is a lot of
string processing. I have another function "findVarName" that returns the
variable string required for a given variable and model type. I often call
this function like this: f->$findVarName()$.
I am including these extra scripts so you can take a look at what I am
doing.
https://github.com/ghassan-alaka/GPLOT/blob/master/ncl/GPLOT_util.ncl
https://github.com/ghassan-alaka/GPLOT/blob/master/ncl/GPLOT_func.ncl
Best,
Gus
--
Ghassan J. Alaka, Jr., Ph.D.
NOAA/AOML/Hurricane Research Division
4301 Rickenbacker Causeway
Miami. FL 33149
Tel: (305)361-4409
Fax: (305)361-4402
**Opinions expressed in this email are my own and do not represent AOML or
NOAA.**
On Thu, Dec 12, 2019 at 2:56 PM Dennis Shea <shea at ucar.edu> wrote:
> I have Ghasson's test script on 3 different systems with no issues.
>
> ---
> MAC:
> %> uname -a
> Darwin sherkin 18.7.0 Darwin Kernel Version 18.7.0: Tue Aug 20 16:57:14
> PDT 2019; root:xnu-4903.271.2~2/RELEASE_X86_64 x86_64
>
> %> gcc --version
> gcc (GCC) 6.2.0
> ---
> %> uname -a
> Linux andre.cgd.ucar.edu 3.10.0-1062.4.1.el7.x86_64 #1 SMP Fri Oct 18
> 17:15:30 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux
>
> %> gcc --version
> gcc (GCC) 4.8.5 20150623 (Red Hat 4.8.5-39)
> ---
> %>uname -a
> Linux cheyenne3 4.12.14-94.41-default #1 SMP Wed Oct 31 12:25:04 UTC 2018
> (3090901) x86_64 x86_64 x86_64 GNU/Linux
>
> %>gcc --version
> gcc (SUSE Linux) 4.8.5
> =====
>
> Comment: I always recommend prototyping function/procedure arguments and
> using 'local'.
>
> Using a bogus example based on Rick's trivial example:
>
> function foo(f[1]:file) ; argument prototype [scalar] and must be type
> 'file'
> local v , x,y,z ; local variables
> begin
> v = f->$VAR$
>
> x = 5*v
> y = sqrt(v)
> z = x/y
> return v
> end
>
> Upon encountering th 'end', the 'local' x,y,z are automatically deleted.
> Also,, I think, a copy of local 'v' is is returned and local v is deleted.
>
> Cheers
> D
>
> On Thu, Dec 12, 2019 at 10:29 AM Rick Brownrigg <brownrig at ucar.edu> wrote:
>
>> Hi,
>>
>> For better or worse, I am unable to duplicate this (with ncl v6.6.2). I
>> didn't have access to the two GPLOT scripts you load at the top of your
>> script, so I don't know what your function is doing. I cobbled up a simple
>> one like:
>>
>> VAR = "UGRD_P0_L100_GLL0"
>> ...
>> function foo(f)
>> begin
>> v = f->$VAR$
>> return v
>> end
>> ...
>> v = foo(f)
>> ...print memory stats...
>> delete([/f,v/])
>> ...print memory stats...
>>
>> So, I'm not sure what to recommend here. By chance, is there a lot of
>> string processing going on somewhere?
>>
>> Rick
>> On Thu, Dec 12, 2019 at 9:01 AM Ghassan Alaka - NOAA Federal <
>> ghassan.alaka at noaa.gov> wrote:
>>
>>> Here you go:
>>>
>>> natl00l.2019082700.hafsprs.synoptic.0p03.f000.grb2
>>> <https://drive.google.com/a/noaa.gov/file/d/1eyEyz1AlbpVBn8Acv7i1Mdea5vXILP5I/view?usp=drive_web>
>>>
>>>
>>> Best,
>>> Gus
>>>
>>> --
>>> Ghassan J. Alaka, Jr., Ph.D.
>>> NOAA/AOML/Hurricane Research Division
>>> 4301 Rickenbacker Causeway
>>> Miami. FL 33149
>>> Tel: (305)361-4409
>>> Fax: (305)361-4402
>>>
>>> **Opinions expressed in this email are my own and do not represent AOML
>>> or NOAA.**
>>>
>>>
>>>
>>> On Thu, Dec 12, 2019 at 10:19 AM Dennis Shea <shea at ucar.edu> wrote:
>>>
>>>> Google drive should beok.
>>>>
>>>> THX
>>>> D
>>>>
>>>> On Thu, Dec 12, 2019 at 8:06 AM Ghassan Alaka - NOAA Federal <
>>>> ghassan.alaka at noaa.gov> wrote:
>>>>
>>>>> I'm having trouble using the FTP option because I am behind a NOAA
>>>>> firewall. Do you by any chance have access to NOAA's Jet supercomputer?
>>>>>
>>>>> Can I share it with you via Google Drive? The file is 1.8GB.
>>>>>
>>>>> Best,
>>>>> Gus
>>>>>
>>>>> --
>>>>> Ghassan J. Alaka, Jr., Ph.D.
>>>>> NOAA/AOML/Hurricane Research Division
>>>>> 4301 Rickenbacker Causeway
>>>>> Miami. FL 33149
>>>>> Tel: (305)361-4409
>>>>> Fax: (305)361-4402
>>>>>
>>>>> **Opinions expressed in this email are my own and do not represent
>>>>> AOML or NOAA.**
>>>>>
>>>>>
>>>>>
>>>>> On Thu, Dec 12, 2019 at 9:57 AM Dennis Shea <shea at ucar.edu> wrote:
>>>>>
>>>>>> Hello,
>>>>>>
>>>>>> Did u send the file?
>>>>>> We can not look at ftp for security reasons.
>>>>>>
>>>>>> Cheers
>>>>>> D
>>>>>>
>>>>>> On Wed, Dec 11, 2019 at 3:10 PM Dennis Shea <shea at ucar.edu> wrote:
>>>>>>
>>>>>>> Ummm, we can look.
>>>>>>>
>>>>>>> offline
>>>>>>>
>>>>>>> Please make "natl00l.2019082700.hafsprs.synoptic.0p03.f000.grb2"
>>>>>>> available.
>>>>>>> Dropbox
>>>>>>>
>>>>>>> or ftp
>>>>>>>
>>>>>>> ftp ftp.cgd.ucar.edu
>>>>>>> anonymous
>>>>>>> your_email
>>>>>>> cd incoming
>>>>>>> put natl00l.2019082700.hafsprs.synoptic.0p03.f000.grb2
>>>>>>> quit
>>>>>>>
>>>>>>> After successful transfer, please send us an email.
>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> On Wed, Dec 11, 2019 at 3:00 PM Ghassan Alaka - NOAA Federal via
>>>>>>> ncl-talk <ncl-talk at ucar.edu> wrote:
>>>>>>>
>>>>>>>> Hi All,
>>>>>>>>
>>>>>>>> Similar but different to a previous submission (see link below), I
>>>>>>>> am experiencing a memory leak when reading data from a file that was read
>>>>>>>> in with the function "addfile". There are some key differences in my
>>>>>>>> situation from the one in the link:
>>>>>>>> 1) I am creating variable "f" using "f = addfile()" in my parent
>>>>>>>> script to read a GRIB2 data file.
>>>>>>>> 2) I wrote a custom function that resides in another script and
>>>>>>>> takes "f" as an input variable. This was done because an undetermined
>>>>>>>> number of variables need to be read from the GRIB2 data file described by
>>>>>>>> "f". Plus, a lot of ugly logic related to the actual reading of the data is
>>>>>>>> stored in this function as well.
>>>>>>>> 3) In the function, I read the data into a variable "V" and return
>>>>>>>> that variable to the parent script.
>>>>>>>> 4) In the parent script, a loop iterates over multiple GRIB2 data
>>>>>>>> files. The variables "f" and "V" are eventually overwritten by new data.
>>>>>>>> 5) I am very careful to delete old variables, including "f" and "V"
>>>>>>>> in the parent script.
>>>>>>>> 6) I have monitored the memory increasing until it sometimes
>>>>>>>> reaches the maximum value and the system kills my job. This out-of-memory
>>>>>>>> kill occurs after several iterations through the list of GRIB2 files.
>>>>>>>>
>>>>>>>> http://mailman.ucar.edu/pipermail/ncl-talk/2017-July/009512.html
>>>>>>>>
>>>>>>>> Given this setup, I am experiencing a memory leak when reading the
>>>>>>>> data from file "f" into variable "V" in the function called from the parent
>>>>>>>> script. I can't delete "f" within the function because it is an input
>>>>>>>> argument and its deletion is not allowed. I directly tested the following
>>>>>>>> and monitored memory usage with "systemfunc("ps u")" calls from within the
>>>>>>>> script.
>>>>>>>> a) If I read the data from file "f" into variable "V" in the
>>>>>>>> primary script, then "delete([/V,f/])", there is no apparent memory leak.
>>>>>>>> b) If I create a variable "V" using the "new()" function with the
>>>>>>>> same dimensions as the data from the GRIB2 file, then delete it, there is
>>>>>>>> no apparent memory leak.
>>>>>>>> c) If I create file "f" from the primary script, then pass "f" as a
>>>>>>>> function argument and read the data into variable "V", then return "V" to
>>>>>>>> the primary script, then "delete([/V,f/])", *there is an apparent
>>>>>>>> memory leak*.
>>>>>>>>
>>>>>>>> I have not tested what happens if file "f" is defined as a local
>>>>>>>> variable in the function and then explicitly deleted before "V" is returned
>>>>>>>> to the parent script. Although that may work, it is not a feasible solution
>>>>>>>> given my the script would slow down significantly by repeatedly executing
>>>>>>>> "f=addfile()". Another possible solution is to execute "f->VAR" from within
>>>>>>>> the primary script. However, such a change would negatively impact these
>>>>>>>> scripts since I have a lot of logic related to the creation of "V" that I
>>>>>>>> am able to hide from the primary script.
>>>>>>>>
>>>>>>>> Below I pasted a sample script using NCL V6.6.2 and its output that
>>>>>>>> I used to test these issues. The issue was reproducible using NCL V6.5.0.
>>>>>>>> Can variable type "file" be safely passed as a function argument? Do you
>>>>>>>> think the reference to "f" is being copied by the function and the extra
>>>>>>>> memory associated with the "f->" statement is never released? Any ideas on
>>>>>>>> the best way forward while maintaining the efficiency of my function usage?
>>>>>>>>
>>>>>>>> Best,
>>>>>>>> Gus
>>>>>>>>
>>>>>>>> --
>>>>>>>> Ghassan J. Alaka, Jr., Ph.D.
>>>>>>>> NOAA/AOML/Hurricane Research Division
>>>>>>>> 4301 Rickenbacker Causeway
>>>>>>>> Miami. FL 33149
>>>>>>>> Tel: (305)361-4409
>>>>>>>> Fax: (305)361-4402
>>>>>>>>
>>>>>>>> **Opinions expressed in this email are my own and do not represent
>>>>>>>> AOML or NOAA.**
>>>>>>>>
>>>>>>>>
>>>>>>>> *NCL SCRIPT*
>>>>>>>> ; A simple NCL script that shows a potential memory leak when using
>>>>>>>> "addfile"
>>>>>>>>
>>>>>>>>
>>>>>>>> load "$GPLOT_DIR/ncl/GPLOT_util.ncl"
>>>>>>>> load "$GPLOT_DIR/ncl/GPLOT_func.ncl"
>>>>>>>>
>>>>>>>> begin
>>>>>>>>
>>>>>>>>
>>>>>>>> IFILE =
>>>>>>>> "/PATH/TO/FILE/natl00l.2019082700.hafsprs.synoptic.0p03.f000.grb2"
>>>>>>>> VAR = "UGRD_P0_L100_GLL0"
>>>>>>>> ILON = "lon_0"
>>>>>>>>
>>>>>>>> print("Checking original memory usage")
>>>>>>>> psrep = systemfunc("ps u")
>>>>>>>> print("MEMORY CHECK 01: "+psrep)
>>>>>>>>
>>>>>>>> print("")
>>>>>>>> print("Using addfile to read the data file")
>>>>>>>> f = addfile(IFILE,"r")
>>>>>>>> psrep = systemfunc("ps u")
>>>>>>>> print("MEMORY CHECK 02: "+psrep)
>>>>>>>>
>>>>>>>> print("")
>>>>>>>> print("Creating new variable V (from data file)")
>>>>>>>> V = f->$VAR$
>>>>>>>> psrep = systemfunc("ps u")
>>>>>>>> print("MEMORY CHECK 03: "+psrep)
>>>>>>>>
>>>>>>>> print("")
>>>>>>>> print("Deleting variable V (from data file)")
>>>>>>>> ;delete([/V,f/])
>>>>>>>> delete(V)
>>>>>>>> psrep = systemfunc("ps u")
>>>>>>>> print("MEMORY CHECK 04: "+psrep)
>>>>>>>>
>>>>>>>> print("")
>>>>>>>> print("Deleting variable f (data file)")
>>>>>>>> ;delete([/V,f/])
>>>>>>>> delete(f)
>>>>>>>> psrep = systemfunc("ps u")
>>>>>>>> print("MEMORY CHECK 05: "+psrep)
>>>>>>>>
>>>>>>>> print("")
>>>>>>>> print("Creating new variable V (not from data file)")
>>>>>>>> V := new((/46,2001,3501/),"float")
>>>>>>>> psrep = systemfunc("ps u")
>>>>>>>> print("MEMORY CHECK 06: "+psrep)
>>>>>>>>
>>>>>>>> print("")
>>>>>>>> print("Deleting variable V (not from data file)")
>>>>>>>> delete(V)
>>>>>>>> psrep = systemfunc("ps u")
>>>>>>>> print("MEMORY CHECK 07: "+psrep)
>>>>>>>>
>>>>>>>> print("")
>>>>>>>> print("Creating new variable V2 (not from data file)")
>>>>>>>> V2 = new((/46,2001,3501/),"float")
>>>>>>>> psrep = systemfunc("ps u")
>>>>>>>> print("MEMORY CHECK 08: "+psrep)
>>>>>>>>
>>>>>>>> print("")
>>>>>>>> print("Deleting variable V2 (not from data file)")
>>>>>>>> delete(V2)
>>>>>>>> psrep = systemfunc("ps u")
>>>>>>>> print("MEMORY CHECK 09: "+psrep)
>>>>>>>>
>>>>>>>> list_vars()
>>>>>>>>
>>>>>>>>
>>>>>>>> print("")
>>>>>>>> print("Creating new variable V (from data file via function)")
>>>>>>>> f = addfile(IFILE,"r")
>>>>>>>> LON = f->$ILON$-360.
>>>>>>>> LON!0 = "lon"
>>>>>>>> LON&lon = LON
>>>>>>>> Vdims :=
>>>>>>>> dimsizes(getfilevardimsizes(f,findVarName("HAFS","U","850")))
>>>>>>>> VdNames := getfilevardims(f,findVarName("HAFS","U","850"))
>>>>>>>> Vdims at dNames = VdNames
>>>>>>>> V :=
>>>>>>>> getVar2d(f,"HAFS",(/"U","850","0"/),Vdims,(/40.,0.,-100.,-20./),1.,False,LON)
>>>>>>>> psrep = systemfunc("ps u")
>>>>>>>> print("MEMORY CHECK 10: "+psrep)
>>>>>>>>
>>>>>>>> print("")
>>>>>>>> print("Deleting variable V (from data file via function)")
>>>>>>>> delete([/V,f,Vdims,VdNames,LON/])
>>>>>>>> psrep = systemfunc("ps u")
>>>>>>>> print("MEMORY CHECK 11: "+psrep)
>>>>>>>>
>>>>>>>> list_vars()
>>>>>>>>
>>>>>>>> end
>>>>>>>>
>>>>>>>>
>>>>>>>> *OUTPUT*
>>>>>>>> [Ghassan.Alaka at v139 ~]$ ncl ~/NCL/scripts/memoryleak.ncl
>>>>>>>> Copyright (C) 1995-2019 - All Rights Reserved
>>>>>>>> University Corporation for Atmospheric Research
>>>>>>>> NCAR Command Language Version 6.6.2
>>>>>>>> The use of this software is governed by a License Agreement.
>>>>>>>> See http://www.ncl.ucar.edu/ for more details.
>>>>>>>> (0) Checking original memory usage
>>>>>>>> (0) MEMORY CHECK 01: USER PID %CPU %MEM VSZ RSS TTY
>>>>>>>> STAT START TIME COMMAND
>>>>>>>> (1) MEMORY CHECK 01: Ghassan+ 7761 0.0 0.0 119692 2116
>>>>>>>> pts/0 Ss 20:46 0:00 -bin/tcsh
>>>>>>>> (2) MEMORY CHECK 01: Ghassan+ 8129 33.0 0.1 232392 126320
>>>>>>>> pts/0 S+ 20:50 0:00 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (3) MEMORY CHECK 01: Ghassan+ 8134 0.0 0.1 232392 122128
>>>>>>>> pts/0 S+ 20:50 0:00 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (4) MEMORY CHECK 01: Ghassan+ 8135 0.0 0.0 155372 1860
>>>>>>>> pts/0 R+ 20:50 0:00 ps u
>>>>>>>> (0)
>>>>>>>> (0) Using addfile to read the data file
>>>>>>>> (0) MEMORY CHECK 02: USER PID %CPU %MEM VSZ RSS TTY
>>>>>>>> STAT START TIME COMMAND
>>>>>>>> (1) MEMORY CHECK 02: Ghassan+ 7761 0.0 0.0 119692 2116
>>>>>>>> pts/0 Ss 20:46 0:00 -bin/tcsh
>>>>>>>> (2) MEMORY CHECK 02: Ghassan+ 8129 38.0 0.2 241416 135612
>>>>>>>> pts/0 S+ 20:50 0:00 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (3) MEMORY CHECK 02: Ghassan+ 8137 0.0 0.1 241416 131100
>>>>>>>> pts/0 S+ 20:50 0:00 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (4) MEMORY CHECK 02: Ghassan+ 8138 0.0 0.0 155372 1856
>>>>>>>> pts/0 R+ 20:50 0:00 ps u
>>>>>>>> (0)
>>>>>>>> (0) Creating new variable V (from data file)
>>>>>>>> (0) MEMORY CHECK 03: USER PID %CPU %MEM VSZ RSS TTY
>>>>>>>> STAT START TIME COMMAND
>>>>>>>> (1) MEMORY CHECK 03: Ghassan+ 7761 0.0 0.0 119692 2116
>>>>>>>> pts/0 Ss 20:46 0:00 -bin/tcsh
>>>>>>>> (2) MEMORY CHECK 03: Ghassan+ 8129 95.4 2.2 1597980 1492360
>>>>>>>> pts/0 S+ 20:50 0:14 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (3) MEMORY CHECK 03: Ghassan+ 8150 0.0 2.2 1597980 1487716
>>>>>>>> pts/0 S+ 20:50 0:00 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (4) MEMORY CHECK 03: Ghassan+ 8151 0.0 0.0 155372 1856
>>>>>>>> pts/0 R+ 20:50 0:00 ps u
>>>>>>>> (0)
>>>>>>>> (0) Deleting variable V (from data file)
>>>>>>>> (0) MEMORY CHECK 04: USER PID %CPU %MEM VSZ RSS TTY
>>>>>>>> STAT START TIME COMMAND
>>>>>>>> (1) MEMORY CHECK 04: Ghassan+ 7761 0.0 0.0 119692 2116
>>>>>>>> pts/0 Ss 20:46 0:00 -bin/tcsh
>>>>>>>> (2) MEMORY CHECK 04: Ghassan+ 8129 95.5 0.3 339176 233560
>>>>>>>> pts/0 S+ 20:50 0:14 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (3) MEMORY CHECK 04: Ghassan+ 8152 0.0 0.3 339176 228984
>>>>>>>> pts/0 S+ 20:50 0:00 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (4) MEMORY CHECK 04: Ghassan+ 8153 0.0 0.0 155372 1864
>>>>>>>> pts/0 R+ 20:50 0:00 ps u
>>>>>>>> (0)
>>>>>>>> (0) Deleting variable f (data file)
>>>>>>>> (0) MEMORY CHECK 05: USER PID %CPU %MEM VSZ RSS TTY
>>>>>>>> STAT START TIME COMMAND
>>>>>>>> (1) MEMORY CHECK 05: Ghassan+ 7761 0.0 0.0 119692 2116
>>>>>>>> pts/0 Ss 20:46 0:00 -bin/tcsh
>>>>>>>> (2) MEMORY CHECK 05: Ghassan+ 8129 95.6 0.1 235108 129496
>>>>>>>> pts/0 S+ 20:50 0:14 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (3) MEMORY CHECK 05: Ghassan+ 8154 0.0 0.1 235108 124916
>>>>>>>> pts/0 S+ 20:50 0:00 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (4) MEMORY CHECK 05: Ghassan+ 8155 0.0 0.0 155372 1860
>>>>>>>> pts/0 R+ 20:50 0:00 ps u
>>>>>>>> (0)
>>>>>>>> (0) Creating new variable V (not from data file)
>>>>>>>> (0) MEMORY CHECK 06: USER PID %CPU %MEM VSZ RSS TTY
>>>>>>>> STAT START TIME COMMAND
>>>>>>>> (1) MEMORY CHECK 06: Ghassan+ 7761 0.0 0.0 119692 2116
>>>>>>>> pts/0 Ss 20:46 0:00 -bin/tcsh
>>>>>>>> (2) MEMORY CHECK 06: Ghassan+ 8129 93.3 2.1 1493912 1388304
>>>>>>>> pts/0 S+ 20:50 0:14 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (3) MEMORY CHECK 06: Ghassan+ 8156 0.0 2.0 1493912 1383720
>>>>>>>> pts/0 S+ 20:50 0:00 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (4) MEMORY CHECK 06: Ghassan+ 8157 0.0 0.0 155372 1864
>>>>>>>> pts/0 R+ 20:50 0:00 ps u
>>>>>>>> (0)
>>>>>>>> (0) Deleting variable V (not from data file)
>>>>>>>> (0) MEMORY CHECK 07: USER PID %CPU %MEM VSZ RSS TTY
>>>>>>>> STAT START TIME COMMAND
>>>>>>>> (1) MEMORY CHECK 07: Ghassan+ 7761 0.0 0.0 119692 2116
>>>>>>>> pts/0 Ss 20:46 0:00 -bin/tcsh
>>>>>>>> (2) MEMORY CHECK 07: Ghassan+ 8129 93.3 0.1 235108 129500
>>>>>>>> pts/0 S+ 20:50 0:14 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (3) MEMORY CHECK 07: Ghassan+ 8158 0.0 0.1 235108 124916
>>>>>>>> pts/0 S+ 20:50 0:00 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (4) MEMORY CHECK 07: Ghassan+ 8159 0.0 0.0 155372 1860
>>>>>>>> pts/0 R+ 20:50 0:00 ps u
>>>>>>>> (0)
>>>>>>>> (0) Creating new variable V2 (not from data file)
>>>>>>>> (0) MEMORY CHECK 08: USER PID %CPU %MEM VSZ RSS TTY
>>>>>>>> STAT START TIME COMMAND
>>>>>>>> (1) MEMORY CHECK 08: Ghassan+ 7761 0.0 0.0 119692 2116
>>>>>>>> pts/0 Ss 20:46 0:00 -bin/tcsh
>>>>>>>> (2) MEMORY CHECK 08: Ghassan+ 8129 91.4 2.1 1493912 1388304
>>>>>>>> pts/0 S+ 20:50 0:15 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (3) MEMORY CHECK 08: Ghassan+ 8161 0.0 2.0 1493912 1383720
>>>>>>>> pts/0 S+ 20:51 0:00 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (4) MEMORY CHECK 08: Ghassan+ 8162 0.0 0.0 155372 1860
>>>>>>>> pts/0 R+ 20:51 0:00 ps u
>>>>>>>> (0)
>>>>>>>> (0) Deleting variable V2 (not from data file)
>>>>>>>> (0) MEMORY CHECK 09: USER PID %CPU %MEM VSZ RSS TTY
>>>>>>>> STAT START TIME COMMAND
>>>>>>>> (1) MEMORY CHECK 09: Ghassan+ 7761 0.0 0.0 119692 2116
>>>>>>>> pts/0 Ss 20:46 0:00 -bin/tcsh
>>>>>>>> (2) MEMORY CHECK 09: Ghassan+ 8129 91.4 0.1 235108 129500
>>>>>>>> pts/0 S+ 20:50 0:15 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (3) MEMORY CHECK 09: Ghassan+ 8163 0.0 0.1 235108 124916
>>>>>>>> pts/0 S+ 20:51 0:00 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (4) MEMORY CHECK 09: Ghassan+ 8164 0.0 0.0 155372 1860
>>>>>>>> pts/0 R+ 20:51 0:00 ps u
>>>>>>>>
>>>>>>>> float pi [ 1 ]
>>>>>>>>
>>>>>>>> string GPLOT [ 1 ]
>>>>>>>>
>>>>>>>> float ms2kts [ 1 ]
>>>>>>>>
>>>>>>>> string psrep [ 5 ]
>>>>>>>>
>>>>>>>> float fval3 [ 1 ]
>>>>>>>>
>>>>>>>> integer fval2 [ 1 ]
>>>>>>>>
>>>>>>>> string ILON [ 1 ]
>>>>>>>>
>>>>>>>> float d2r [ 1 ]
>>>>>>>>
>>>>>>>> string VAR [ 1 ]
>>>>>>>>
>>>>>>>> float r [ 1 ]
>>>>>>>>
>>>>>>>> float g [ 1 ]
>>>>>>>>
>>>>>>>> logical GSN_OLD_LABELBAR [ 1 ]
>>>>>>>>
>>>>>>>> string IFILE [ 1 ]
>>>>>>>>
>>>>>>>> float fval [ 1 ]
>>>>>>>>
>>>>>>>> float omega [ 1 ]
>>>>>>>> (0)
>>>>>>>> (0) Creating new variable V (from data file via function)
>>>>>>>> (0) MEMORY CHECK 10: USER PID %CPU %MEM VSZ RSS TTY
>>>>>>>> STAT START TIME COMMAND
>>>>>>>> (1) MEMORY CHECK 10: Ghassan+ 7761 0.0 0.0 119692 2116
>>>>>>>> pts/0 Ss 20:46 0:00 -bin/tcsh
>>>>>>>> (2) MEMORY CHECK 10: Ghassan+ 8129 90.5 0.4 384024 277172
>>>>>>>> pts/0 S+ 20:50 0:16 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (3) MEMORY CHECK 10: Ghassan+ 8166 0.0 0.4 384024 272304
>>>>>>>> pts/0 S+ 20:51 0:00 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (4) MEMORY CHECK 10: Ghassan+ 8167 0.0 0.0 155372 1860
>>>>>>>> pts/0 R+ 20:51 0:00 ps u
>>>>>>>> (0)
>>>>>>>> (0) Deleting variable V (from data file via function)
>>>>>>>> (0) MEMORY CHECK 11: USER PID %CPU %MEM VSZ RSS TTY
>>>>>>>> STAT START TIME COMMAND
>>>>>>>> (1) MEMORY CHECK 11: Ghassan+ 7761 0.0 0.0 119692 2116
>>>>>>>> pts/0 Ss 20:46 0:00 -bin/tcsh
>>>>>>>> (2) MEMORY CHECK 11: Ghassan+ 8129 90.6 0.2 267228 161908
>>>>>>>> pts/0 S+ 20:50 0:16 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (3) MEMORY CHECK 11: Ghassan+ 8168 0.0 0.2 267228 157040
>>>>>>>> pts/0 S+ 20:51 0:00 ncl /home/Ghassan.Alaka/NCL/scripts/memoryleak.ncl
>>>>>>>> (4) MEMORY CHECK 11: Ghassan+ 8169 0.0 0.0 155372 1860
>>>>>>>> pts/0 R+ 20:51 0:00 ps u
>>>>>>>>
>>>>>>>> float pi [ 1 ]
>>>>>>>>
>>>>>>>> string GPLOT [ 1 ]
>>>>>>>>
>>>>>>>> float ms2kts [ 1 ]
>>>>>>>>
>>>>>>>> string psrep [ 5 ]
>>>>>>>>
>>>>>>>> float fval3 [ 1 ]
>>>>>>>>
>>>>>>>> integer fval2 [ 1 ]
>>>>>>>>
>>>>>>>> string ILON [ 1 ]
>>>>>>>>
>>>>>>>> float d2r [ 1 ]
>>>>>>>>
>>>>>>>> string VAR [ 1 ]
>>>>>>>>
>>>>>>>> float r [ 1 ]
>>>>>>>>
>>>>>>>> float g [ 1 ]
>>>>>>>>
>>>>>>>> logical GSN_OLD_LABELBAR [ 1 ]
>>>>>>>>
>>>>>>>> string IFILE [ 1 ]
>>>>>>>>
>>>>>>>> float fval [ 1 ]
>>>>>>>>
>>>>>>>> float omega [ 1 ]
>>>>>>>>
>>>>>>>> _______________________________________________
>>>>>>>> ncl-talk mailing list
>>>>>>>> ncl-talk at ucar.edu
>>>>>>>> List instructions, subscriber options, unsubscribe:
>>>>>>>> http://mailman.ucar.edu/mailman/listinfo/ncl-talk
>>>>>>>
>>>>>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.ucar.edu/pipermail/ncl-talk/attachments/20191213/2b65927c/attachment.html>
More information about the ncl-talk
mailing list