Torque/PBS can not find munge.socket.2

Question

I am trying to create a virtual cluster for my MPI classes so i can work at home and not be at the university labs the whole day. I can not figure out for 2 days now how to fix this problem with munge.

The output of the problem i have is this

[root@localhost lumx]# qmgr -c "set server acl_hosts = mars"
munge_encode failed: Failed to access "/var/run/munge/munge.socket.2": No such file or directory (6)
Unable to communicate with localhost(127.0.0.1)
Communication failure.
qmgr: cannot connect to server  (errno=15009) munge executable not found, unable to authenticate

My hosts file looks like this

127.0.0.1    localhost localhost.localdomain localhost4 localhost4.localdomain4 
::1         localhost localhost.localdomain localhost6 localhost6.localdomain6
127.0.0.1 mpimaster localhost.localdomain localhost

I am tried to read as much as I could and I ended up with these guides,Getting started with Open MPI on Fedora, Installing Torque/PBS job scheduler on Ubuntu 14.04 LTS,TORQUE Arch Linux, http://juanjose.garciaripoll.com/fedora-cluster/5-torque-pbs-queue

score 2 · Answer 1 · answered Sep 17 '16 at 22:34

2

I solved it. I just have to force it to start and use syslogs because there are some permission problems for now.

The command I use is

munged --force --syslog

answered Sep 17 '16 at 22:34

Lum Zhaveli

175
2
18

score 0 · Answer 2 · answered Nov 13 '19 at 08:21

For Opensuse Leap 15.X, you need to start the munge service by

sudo service munge start

sudo service munge status

If the service is shown as active, you can verify it by now typing:

munge -n

That should return some results like below:

MUNGE:AwQFAAAHiPEv+E6Ezy2HVHUwo5PZ2fkbbr4yP7pZZA9Yo6BWQdAFGVRNkhNbRkvd9zNAvnpg0iQzkjg+WW6HdIix48nKrA0QnjispII4RoT1UqZLh7ybIl5/WIvd3ta85v1KV8A=:

Torque/PBS can not find munge.socket.2

2 Answers2