Scheduled Downtime
On Friday 21 April 2023 @ 5pm MT, this website will be down for maintenance and expected to return online the morning of 24 April 2023 at the latest

Segmentation fault - invalid memory reference.

H

Huanyu Ge

New member
Hi,

I am currently using ERA5 data to get parameters such as QCLOUD.

The previous steps (geogrid, ungrib, metgrid, real programs) run successfully. However, when I run wrf.exe, following error occurred.

I have attached my namelist.input, rsl.error.0000 and rsl.out.0000 for review.

Thank you very much!


Program received signal SIGSEGV: Segmentation fault - invalid memory reference.

Backtrace for this error:
#0 0x7f18965dc960 in ???
#1 0x7f18965dbac5 in ???
#2 0x7f18961cc51f in ???
at ./signal/../sysdeps/unix/sysv/linux/x86_64/libc_sigaction.c:0
#3 0x55a79ff9fe20 in ???
#4 0x55a79ffa1977 in ???
#5 0x55a79ffb3b9f in ???
#6 0x55a79ffb65ec in ???
#7 0x55a79f486e52 in ???
#8 0x55a79f5ba3d1 in ???
#9 0x55a79f052b7e in ???
#10 0x55a79ee699ba in ???
#11 0x55a79e4793de in ???
#12 0x55a79e468b77 in ???
#13 0x55a79e467ffe in ???
#14 0x7f18961b3d8f in __libc_start_call_main
at ../sysdeps/nptl/libc_start_call_main.h:58
#15 0x7f18961b3e3f in __libc_start_main_impl
at ../csu/libc-start.c:392
#16 0x55a79e468034 in ???
#17 0xffffffffffffffff in ???
 

Attachments

  • namelist.input
    3.6 KB · Views: 4
  • rsl.error.0000
    17.6 KB · Views: 3
  • rsl.out.0000
    16.5 KB · Views: 0
Huanyu Ge said:
Hi,

I am currently using ERA5 data to get parameters such as QCLOUD.

The previous steps (geogrid, ungrib, metgrid, real programs) run successfully. However, when I run wrf.exe, following error occurred.

I have attached my namelist.input, rsl.error.0000 and rsl.out.0000 for review.

Thank you very much!


Program received signal SIGSEGV: Segmentation fault - invalid memory reference.

Backtrace for this error:
#0 0x7f18965dc960 in ???
#1 0x7f18965dbac5 in ???
#2 0x7f18961cc51f in ???
at ./signal/../sysdeps/unix/sysv/linux/x86_64/libc_sigaction.c:0
#3 0x55a79ff9fe20 in ???
#4 0x55a79ffa1977 in ???
#5 0x55a79ffb3b9f in ???
#6 0x55a79ffb65ec in ???
#7 0x55a79f486e52 in ???
#8 0x55a79f5ba3d1 in ???
#9 0x55a79f052b7e in ???
#10 0x55a79ee699ba in ???
#11 0x55a79e4793de in ???
#12 0x55a79e468b77 in ???
#13 0x55a79e467ffe in ???
#14 0x7f18961b3d8f in __libc_start_call_main
at ../sysdeps/nptl/libc_start_call_main.h:58
#15 0x7f18961b3e3f in __libc_start_main_impl
at ../csu/libc-start.c:392
#16 0x55a79e468034 in ???
#17 0xffffffffffffffff in ???
Click to expand...
Code: 
cu_physics                          = 0,    0,    0,

First thing I see is that you need to have a cumulous scheme for the first domain. Since the grid spacing is above 4km you would need a cumulous scheme to parameterize the clouds.

In the RSL files I see this error

Code: 
d03 2009-04-22_00:00:50            2  points exceeded v_cfl = 2 in domain d03 at time 2009-04-22_00:00:50 hours
d03 2009-04-22_00:00:50 Max   W:     32      2     24 W:  -20.05  w-cfl:    2.05  dETA:    0.03
d03 2009-04-22_00:00:50           61  points exceeded v_cfl = 2 in domain d03 at time 2009-04-22_00:00:50 hours
d03 2009-04-22_00:00:50 Max   W:     31      2     25 W:   34.09  w-cfl:    2.89  dETA:    0.02
d03 2009-04-22_00:00:50          595  points exceeded v_cfl = 2 in domain d03 at time 2009-04-22_00:00:50 hours
d03 2009-04-22_00:00:50 Max   W:      5     50     26 W:  -49.19  w-cfl:    4.77  dETA:    0.02
Timing for main: time 2009-04-22_00:01:40 on domain   3:    0.52128 elapsed seconds
Timing for main: time 2009-04-22_00:01:40 on domain   2:    1.14177 elapsed seconds
Timing for main: time 2009-04-22_00:01:40 on domain   1:    1.79418 elapsed seconds
d02 2009-04-22_00:01:40            4  points exceeded v_cfl = 2 in domain d02 at time 2009-04-22_00:01:40 hours
d02 2009-04-22_00:01:40 Max   W:     41     53     25 W:  -24.30  w-cfl:    2.50  dETA:    0.02
d02 2009-04-22_00:01:40          110  points exceeded v_cfl = 2 in domain d02 at time 2009-04-22_00:01:40 hours
d02 2009-04-22_00:01:40 Max   W:     47     53     23 W:    5.85  w-cfl:    3.46  dETA:    0.03
d02 2009-04-22_00:01:40            2  points exceeded v_cfl = 2 in domain d02 at time 2009-04-22_00:01:40 hours
d02 2009-04-22_00:01:40 Max   W:     43     52     29 W:  -12.24  w-cfl:    2.02  dETA:    0.02
d03 2009-04-22_00:01:40         9337  points exceeded v_cfl = 2 in domain d03 at time 2009-04-22_00:01:40 hours
d03 2009-04-22_00:01:40 Max   W:     31      3     25 W: -232.38  w-cfl:   25.81  dETA:    0.02
d03 2009-04-22_00:01:40        19916  points exceeded v_cfl = 2 in domain d03 at time 2009-04-22_00:01:40 hours
d03 2009-04-22_00:01:40 Max   W:      7     50      9 W: *******  w-cfl: 2060.18  dETA:    0.03
d03 2009-04-22_00:01:40         6187  points exceeded v_cfl = 2 in domain d03 at time 2009-04-22_00:01:40 hours
d03 2009-04-22_00:01:40 Max   W:     19     38     30 W: *******  w-cfl: *******  dETA:    0.02

See this post about cfl errors:
forum.mmm.ucar.edu

Segmentation Faults and CFL Errors

Segmentation faults can be difficult to track down. As there isn't usually a clear error message, it can take some trial and error to figure out the problem. 1) A segmentation fault is often the result of using too many or too few processors, or a bad decomposition. Take a look at this FAQ...
forum.mmm.ucar.edu forum.mmm.ucar.edu

Try turning on w_damping
Code: 
w_damping                           = 0,  -->  w_damping                           = 1,
 
William.Hatheway said:
Code: 
cu_physics                          = 0,    0,    0,

First thing I see is that you need to have a cumulous scheme for the first domain. Since the grid spacing is above 4km you would need a cumulous scheme to parameterize the clouds.

In the RSL files I see this error

Code: 
d03 2009-04-22_00:00:50            2  points exceeded v_cfl = 2 in domain d03 at time 2009-04-22_00:00:50 hours
d03 2009-04-22_00:00:50 Max   W:     32      2     24 W:  -20.05  w-cfl:    2.05  dETA:    0.03
d03 2009-04-22_00:00:50           61  points exceeded v_cfl = 2 in domain d03 at time 2009-04-22_00:00:50 hours
d03 2009-04-22_00:00:50 Max   W:     31      2     25 W:   34.09  w-cfl:    2.89  dETA:    0.02
d03 2009-04-22_00:00:50          595  points exceeded v_cfl = 2 in domain d03 at time 2009-04-22_00:00:50 hours
d03 2009-04-22_00:00:50 Max   W:      5     50     26 W:  -49.19  w-cfl:    4.77  dETA:    0.02
Timing for main: time 2009-04-22_00:01:40 on domain   3:    0.52128 elapsed seconds
Timing for main: time 2009-04-22_00:01:40 on domain   2:    1.14177 elapsed seconds
Timing for main: time 2009-04-22_00:01:40 on domain   1:    1.79418 elapsed seconds
d02 2009-04-22_00:01:40            4  points exceeded v_cfl = 2 in domain d02 at time 2009-04-22_00:01:40 hours
d02 2009-04-22_00:01:40 Max   W:     41     53     25 W:  -24.30  w-cfl:    2.50  dETA:    0.02
d02 2009-04-22_00:01:40          110  points exceeded v_cfl = 2 in domain d02 at time 2009-04-22_00:01:40 hours
d02 2009-04-22_00:01:40 Max   W:     47     53     23 W:    5.85  w-cfl:    3.46  dETA:    0.03
d02 2009-04-22_00:01:40            2  points exceeded v_cfl = 2 in domain d02 at time 2009-04-22_00:01:40 hours
d02 2009-04-22_00:01:40 Max   W:     43     52     29 W:  -12.24  w-cfl:    2.02  dETA:    0.02
d03 2009-04-22_00:01:40         9337  points exceeded v_cfl = 2 in domain d03 at time 2009-04-22_00:01:40 hours
d03 2009-04-22_00:01:40 Max   W:     31      3     25 W: -232.38  w-cfl:   25.81  dETA:    0.02
d03 2009-04-22_00:01:40        19916  points exceeded v_cfl = 2 in domain d03 at time 2009-04-22_00:01:40 hours
d03 2009-04-22_00:01:40 Max   W:      7     50      9 W: *******  w-cfl: 2060.18  dETA:    0.03
d03 2009-04-22_00:01:40         6187  points exceeded v_cfl = 2 in domain d03 at time 2009-04-22_00:01:40 hours
d03 2009-04-22_00:01:40 Max   W:     19     38     30 W: *******  w-cfl: *******  dETA:    0.02

See this post about cfl errors:
forum.mmm.ucar.edu

Segmentation Faults and CFL Errors

Segmentation faults can be difficult to track down. As there isn't usually a clear error message, it can take some trial and error to figure out the problem. 1) A segmentation fault is often the result of using too many or too few processors, or a bad decomposition. Take a look at this FAQ...
forum.mmm.ucar.edu forum.mmm.ucar.edu

Try turning on w_damping
Code: 
w_damping                           = 0,  -->  w_damping                           = 1,
Click to expand...

Thank you very much, William! Your suggestion and the post about cfl errors are very helpful!

I solved this issue by reducing the time step from 60s to 6s, because dx of the smallest domain is only 1.1km. The time step depends on the smallest domain, not the parent domain (dx = 10km).

Now the WRF model can run successfully, and I didn't change the cumulous scheme, I would like to know is the setting for cumulous scheme necessary?

​

 
VFL violation indicates that the model integration is not stable. Decreasing time step and/or increase the value of epssm usually can solve the problem. However, in some cases incorrect input data may also cause such issues.
Please double check your data, and adjust namelist options, then try again.
 
Last edited:
Ming Chen said:
VFL violation indicates that the model integration is not stable. Decreasing time step and/or increase the value of epssm usually can solve the problem. However, in some cases incorrect input data may also cause such an issue.
Please double check your data, and adjust namelist options, then try again.
Click to expand...
Thank you for your reply! I have reduced time step and successfully ran the model around Mt. Zao in Japan. Also, I set w_damping = 1 and smooth_option = 1. Although there are no errors in rsl.error.* files, the simulation results show large fluctuations. I attached plots of LWC, temperature and wind speed varying with time. Do you have any idea why this might be happening?
 

Attachments

  • Smooth-LWC.png
    Smooth-LWC.png
    382.5 KB · Views: 1
  • Smooth-TEMP.png
    Smooth-TEMP.png
    371.4 KB · Views: 1
  • Smooth-V.png
    Smooth-V.png
    387.7 KB · Views: 1
Hi,
These plots show clear diurnal variation, which is physically reasonable. Do you have any observations to compare with the model outputs?
 
Attached figures are extracted from Podolskiy et al. (2012). I focus on WRF simulation for Jizosancho from 4/22/2009 00:00 to 4/24/2009 18:00. It can be observed that the green lines in "TEMP Jizosancho.png" and "V Jizosancho.png" exhibit a clear trend with relatively minor fluctuations. The blue dashed line in "LWC Jizosancho.png" also shows relatively small fluctuations.

I followed the same settings of this paper, such as the Thompson microphysics scheme and other schemes. The time period of my simulation is only from 4/22/2009 00:00 to 4/24/2009 18:00. However, my results fluctuated greatly.
 

Attachments

  • LWC Jizosancho.png
    LWC Jizosancho.png
    340.3 KB · Views: 0
  • V Jizosancho.png
    V Jizosancho.png
    306.1 KB · Views: 0
  • TEMP Jizosancho.png
    TEMP Jizosancho.png
    372.1 KB · Views: 0
You must log in or register to reply here.
Top