nfs version 3 with noac needed

[backelj]

Hello,

We have ABINIT-7.8.2 installed on our system (built with EasyBuild, file ABINIT-7.8.2-ictce-5.5.0.eb). We are using mpiexec to run it as follows:

Code: Select all

mpiexec -pernode -verbose abinit  < Bi2Te2Se_Positron_GGA_SO.files >& out_Positron

It is running fine on our GPFS filesystems mounted using the GPFS clients. But it is not running on our cNFS mounted filesystems. Here it gives this error:

Code: Select all

> cat out_Positron
mpiexec: resolve_exe: found "PATH_TO/bin/abinit" in path.
node  0: name THISNODE, cpu avail 1
mpiexec: process_start_event: evt 2 task 0 on THISNODE.
mpiexec: All 1 task (spawn 0) started.
mpiexec: wait_tasks: waiting for THISNODE.
mpiexec: accept_pmi_conn: cmd=initack pmiid=0.
mpiexec: accept_pmi_conn: rank 0 (spawn 0) checks in.
mpiexec: accept_pmi_conn: cmd=init pmi_version=1 pmi_subversion=1.
File locking failed in ADIOI_Set_lock(fd 4,cmd F_SETLKW/7,type F_RDLCK/0,whence 0) with return value FFFFFFFF and errno 25.
- If the file system is NFS, you need to use NFS version 3, ensure that the lockd daemon is running on all the machines, and mount the directory with the 'noac' option (no attribute caching).
- If the file system is LUSTRE, ensure that the directory is mounted with the 'flock' option.
ADIOI_Set_lock:: No locks available
ADIOI_Set_lock:offset 20, length 20
application called MPI_Abort(MPI_COMM_WORLD, 1) - process 0
mpiexec: process_obit_event: evt 3 task 0 on THISNODE stat 1.
mpiexec: Warning: task 0 exited with status 1. 

So it is complaining on the fact that I do not have the 'noac' option specified in our NFS version 3 mounts. However, we do have this:

Code: Select all

> ssh THISNODE
[root@THISNODE~]# nfsstat -m
/CNFS-MOUNT from GPFS-SERVER:/CNFS-MOUNT
 Flags:   rw,sync,vers=3,rsize=1048576,wsize=1048576,acregmin=0,acregmax=0,acdirmin=0,acdirmax=0,hard,intr,noac,proto=tcp,timeo=600,retrans=2,sec=sys,addr=GPFS-SERVER

As said, it is working fine on our GPFS-mounted filesystems, e.g.:

Code: Select all

GPFS-SERVER:/CNFS-MOUNT on /CNFS-MOUNT type nfs (rw,bg,nfsvers=3,proto=tcp,hard,intr,noac,addr=XXX)
/dev/GPFS-MOUNT on /GPFS-MOUNT type gpfs (rw,mtime,quota=userquota;groupquota;filesetquota,dev=scratch,autostart)

Does anyone have an idea what is wrong here?

Many thanks for your reply, Franky

[Jordan]

Hi,

Could you give more details on how you/someone compiled abinit, version of libraries, NetCDF, ETSF_IO, ... and on your system ?

Have you tried an other MPI code with the same compilation option,version ?

Cheers,

Jordan

[pouillon]

This issue has nothing to do with Abinit but is due to a misconfiguration of your computer. Please ask your system administrator to:

• review and fix the mount parameters of the file system on all nodes;
  
• check that all required daemons are running on all nodes;
  
• provide you with a MPI implementation compatible with your hardware and compiled adequately.