Install and configure SLURM in VMware + Ubuntu

VMware is a common Virtual Machine Software. Installing Ubuntu is also very easy. Migration of vmwarevm many virtual machine software support reuse of virtual machine files elsewhere. For example, the Virtual Machine of VirtualBox and ParalellDesktop itself is a file, just copy it directly. However, by default, VMware creates a virtual machine that contains a pile of files in a folder. First, export the VM to a file. File-& gt; ExporttoOVF.

VMware is a common Virtual Machine Software. Installing Ubuntu is also very easy.

Migration of vmwarevm

Many Virtual Machine Software Support reusing virtual machine files elsewhere. For example, the Virtual Machine of VirtualBox and ParalellDesktop itself is a file, just copy it directly. However, by default, VMware creates a virtual machine that contains a pile of files in a folder. First, export the VM to a file. File-> Export to OVF. Then several files are generated, including a small OVF file and a large virtual disk file. There is also a configuration file... You can copy and paste them in a folder. When importing a VM, double-click the OVF file.

Configure the ssh environment in Ubuntu

The ssh client is installed by default, but the server may need to be manually installed

Sudo apt-get openssh-server

View the information after completion

Ps-e | grep ssh

Check whether the ssh-agent and sshd appear. If it appears, it indicates that the startup is successful and other nodes can access this ubuntu.

Install MUNGE in Ubuntu

SLURM needs a plug-in for security management. MUNGE is one of them. The installation and configuration are simple and simple, but my English is not good, so I was wrong when reading the guide)

MUNGE Installation Guide

In short, install the folder first, and then change the permissions of each folder as required. Change the owner of the folder to the user who uses munge. The folder generated during the MUNGE installation belongs to munge. For example, if my user name is tomxice, You need to modify the user name.

Sudo chmod 0700/etc/munge

Sudo chown tomxice/etc/munge

The command to start munge isMungedNoMunge.

Installing SLURM in Ubuntu

Ubuntu has a package for this software. CallSlurm-llnlThe package named slurm is a network monitoring software, which is not the same as the SLURM discussed here.

However, after I installed it with apt-get, I don't know where his configuration file is. So I compiled the source code. Follow the tutorials on the SLURM official website to complete step by step.

SLURM Installation Guide

The configuration file should be written in detail. A configuration.html file can help you configure the file. If you compile the file from the source code, the file is in the doc directory of the source file. Note that SlurmUser and SlurmdUser are two different variables. The default value is root. If you want to run them with the tomxice user, you must have these two lines.

SlurmUser = tomxice

SlurmdUser = tomxice

Finally, the test run is completed. The result is a BUG at the beginning.

Slurmctld: slurmctld version 2.4.2 started on cluster
Slurmctld: debug3: Trying to load plugin/usr/local/lib/slurm/crypto_munge.so
Slurmctld: debug4:/usr/local/lib/slurm/crypto_munge.so: Does not exist or not a regular file.
Slurmctld: error: Couldn't find the specified plugin name for crypto/munge looking at all files
Slurmctld: error: can't find a plugin for type crypto/munge
Slurmctld: error: cannot resolve crypto plugin operations
Slurmctld: fatal: slurm_cred_creator_ctx_create (null): Operation not permitted
The file crypto_munge.so I found one in/usr/lib/slurm. After copying it, I was prompted that a file named auth_munge.so was missing and I copied it again, so both slurmctl and slurmd can run.
In another window, execute squeue and scontrol commands, but run the sample program.
Srun-N1-l/bin/hostname
The following error occurs:
Tomxice @ master:/home $ srun-N1-l/bin/hostname
Srun: symbol lookup error:/usr/local/lib/slurm/auth_munge.so: undefined symbol: slurm_auth_get_arg_desc

I searched on google and found that some people encountered such problems, but did not say how to solve them.

 

I thought it was a problem with the ubuntu system, but I thought it should not be the case. This is because the Link Library is incorrect. Because there are many files of different quality with the same name in/usr/local/lib/slurm and/usr/lib/slurm, copying is just a deception. So I had to send an email to the senior student. The senior student said that he had forgotten it for a long time, so I had to continue to work. Today, things have turned around. Today, when the GFW network opened, I got to google group. I found the same problem here. I have a reply from Daniel.

On Wed, 21 Sep 2011 00:49:18-0700, Daniel Espling Wrote:
> Hi


> I'm trying to build and install slurm 2.3.0-rc2 on Ubuntu Lucid 10.4,
> But for some reason the crypto/munge plugin is not copied as a part
> The "make install" process:
[Snip]
> Any ideas? I 've confirmed that slurm. conf is in the expected directory
> (/Usr/local/etc), that it's set to use crypto/munge and that munge is
> Installed and running.


Did you have munge development files installed at the time
Configure/make? (I. e. munge-dev or munge-devel package installed)


Check config. log with the following and see if there was any problem
Detecting the munge installation:


  Grep-A3 'checking for munge' config. log


Mark

 

So let's just look at the grep. I found that the slurm did not find munge during installation, but I did install it. It should be like Daniel said, there is also a development library and so on, and then found on this website.

Sudo apt-get install libmunge-dev

Just reinstall slurm.

Tomxice @ master :~ /Downloads/slurm-2.4.2 $ srun/bin/hostname
Master

The installation of SLURM has come to an end.