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

error while running wrf "Segmentation fault - invalid memory reference"

DeebakVijay

New member
Hello,
As a part of my work I am running wrf on a thunderstorm event with different pbl schemes. I have attached the namelist.input , wrfbdy , wrfinput , wrfflowinp , rsl.error , rsl.out files (Please find the attached link)" WRF - Google Drive ". I used 8 cores . The error is in rsl.error.002 and rsl.error.003.
The error is;
Program received signal SIGSEGV: Segmentation fault - invalid memory reference.

Backtrace for this error:
#0 0x2b3cfdb0a3ff in ???
#1 0x27ac0aa in ???
#2 0x27c507c in ???
#3 0x218feeb in ???
#4 0x1a0a8f1 in ???
#5 0x147c72d in ???
#6 0x132681a in ???
#7 0x47d9a1 in ???
#8 0x406403 in ???
#9 0x405dbc in ???
#10 0x2b3cfdaf6554 in ???
#11 0x405df3 in ???
#12 0xffffffffffffffff in ???

  • I tried different no of cores (25 , 36 , 50 , 64) : Produced similar errors at different rsl.error files.
  • I have gone through this " How many processors should I use to run WRF? " ,couldn't follow it .
  • As my work involves comparing different pbl schemes for same simulation (4 wrf runs already successful) , I cannot change much in namelist.input.
  • When the previous wrf runs (with different pbl schemes ; every thing else is same ) encountereed same error , changing the number of cores worked (after lot of trial and error). Also I had to use different number of cores for different schemes , for successful wrf run.
1) How can I rectify this problem?
2)If the issue is with number of cores , how do I decide the number of cores?(Identical wrf runs with just different pbl schemes didn't run successfully with same number of cores).

Hoping for a reply.
Thank you.
Regards,
Deebak Vijay
 
There are a few issues regarding your namlist.input. Please change the following options as below:
dx = 7500, 2500
dy = 7500, 2500
time_step = 45,
radt =15, 15

Then try again.

Also, Please run this case using 25 processors.
 
Top