ZHEGV failed, returncode = 8 4 98

Problems running VASP: crashes, internal errors, "wrong" results.


Moderators: Global Moderator, Moderator

Post Reply
Message
Author
ccccc
Jr. Member
Jr. Member
Posts: 69
Joined: Wed Feb 24, 2010 1:56 pm
License Nr.: 4-1116 5-3053

ZHEGV failed, returncode = 8 4 98

#1 Post by ccccc » Tue Jul 23, 2013 4:34 pm

I have these errors in one of my calculations.

WARNING in EDDRMM: call to ZHEGV failed, returncode = 6 3 66
WARNING in EDDRMM: call to ZHEGV failed, returncode = 8 4 98

and at the end of log file it is written that;

=====================================================================================
= BAD TERMINATION OF ONE OF YOUR APPLICATION PROCESSES
= EXIT CODE: 15
= CLEANING UP REMAINING PROCESSES
= YOU CAN IGNORE THE BELOW CLEANUP MESSAGES
=====================================================================================
APPLICATION TERMINATED WITH THE EXIT STRING: Terminated (signal 15

I wrote this before but got no answer.

When I do the same calculation with same INCAR, but for less number of atoms it works well. When i increase the atom numver above 50, i have this error. Is it because of atom numbers?

thanks
Last edited by ccccc on Tue Jul 23, 2013 4:34 pm, edited 1 time in total.

vanderro
Newbie
Newbie
Posts: 1
Joined: Thu Aug 01, 2013 8:10 pm
License Nr.: 1140

ZHEGV failed, returncode = 8 4 98

#2 Post by vanderro » Thu Aug 01, 2013 8:24 pm

I've received the same error and just saw a post from two years ago with similar problems never really solved.

In my case with vasp 4.6 compiled with mpich 3.0.4 the problem only shows up when running in parallel.
Reducing the number of cores used made the problem go away on mine and in the old post I read here...
But here is the solution that makes me feel better.
I only get the error when I compiled mpich using --enable-threads (defaulted to POSIX)
When I use mpich without threading enabled I have not encountered the problem (on another identical system).

This all happened today so it may not be a final solution but hopefully this may help you.

Currently it is running a large >300 atom system and has not yet made it far enough along to make the complaint if it is related to the system size.
But a diamond tutorial run no longer has the error.
Last edited by vanderro on Thu Aug 01, 2013 8:24 pm, edited 1 time in total.

Post Reply