Installing gridengine in centos 7

zillur · 12-30-2016, 05:55 PM

Hi there,
I was trying to install gridengine in my centos 7 server. I have downloaded all the rpm files and run

Code:

yum install *.rpm

Code:

gridengine-8.1.8-1.el6.x86_64.rpm            gridengine-execd-8.1.8-1.el6.x86_64.rpm    gridengine-qmon-8.1.8-1.el6.x86_64.rpm
gridengine-debuginfo-8.1.8-1.el6.x86_64.rpm  gridengine-qmaster-8.1.8-1.el6.x86_64.rpm

Then I installed the gridengine running sripts at /opt/sge ;

Code:

install_execd  install_qmaster  inst_sge

And finally got:

Code:

/home/zillur/Desktop/zillur/phd/gridengine/sge
[zillur@genomics sge]$ ps -ef | grep sge
zillur   17809  8994  0 19:50 pts/45   00:00:00 grep --color=auto sge
zillur   29491     1  0 Dec29 ?        00:00:29 /opt/sge/bin/lx-amd64/sge_qmaster
zillur   30216     1  0 Dec29 ?        00:00:20 /opt/sge/bin/lx-amd64/sge_execd

But if I run qsub:

Code:

bash: qsub: command not found...

If I want to restart the daemon:

Code:

[zillur@genomics sge]$ service sgeexec.p6444 stop && sgeexec.p6444 start
Redirecting to /bin/systemctl stop  sgeexec.p6444.service
Failed to stop sgeexec.p6444.service: Unit sgeexec.p6444.service not loaded.

Any help with the installation?

Best Regards
Zillur

szboardstretcher · 12-30-2016, 06:22 PM

What do you get if you:

Code:

cd /opt/sge
find . -name qsub*
systemctl list-unit-files | grep sge

zillur · 12-30-2016, 06:40 PM

Thank you very much for your reply. Here is the output:

Code:

[zillur@genomics ~]$ cd /opt/sge/
[zillur@genomics sge]$ find . -name qsub*
./man/man1/qsub.1
./bin/lx-amd64/qsub
[zillur@genomics sge]$ systemctl list-unit-files | grep sge
[zillur@genomics sge]$

The last command didn't give anything. Thanks again.

Best Regards
Zillur

szboardstretcher · 12-30-2016, 07:29 PM

Ok. Lets find the name of the service file that was supposed to be installed and its filename.

Code:

for p in $(rpm -qa | grep gridengine); do
rpm -ql $p | grep service
done

zillur · 12-30-2016, 08:17 PM

Thank you very much for your reply. Here it is:

Code:

[zillur@genomics sge]$ for p in $(rpm -qa | grep gridengine); do
> rpm -ql $p | grep service
> done
/opt/sge/util/rctemplates/systemd/sgeexecd.service
/opt/sge/util/rctemplates/systemd/sgemaster.service
/opt/sge/util/rctemplates/systemd/sgeshadowd.service

Best Regards
Zillur

szboardstretcher · 12-30-2016, 09:03 PM

Ok so lets add /opt/sge/bin/lx-amd64/ to the system path,. you can do this by user(add to .bash_profile) or system wide(add to /etc/environment).

Code:

echo "PATH=$PATH:/opt/sge/bin/lx-amd64/" >> /etc/environment
or
echo "PATH=$PATH:/opt/sge/bin/lx-amd64/" >> /home/YOURUSER/.bash_profile

Log out, log back in. 'qsub' should work as a command now.

And then use the correct systemd service to start, which looks to me like sgemaster, but might be one of the others according to your output above:

Code:

systemctl start sgemaster
systemctl status sgemaster

zillur · 12-30-2016, 11:25 PM

Thank you very much. Here is current situation:

Code:

[zillur@genomics ~]$ echo "PATH=$PATH;/opt/sge/bin/lx-amd64/" >> /home/zillur/.bash_profile
[zillur@genomics ~]$ logout 
Connection to 136.145.231.38 closed.
MacBook-Pro-55:final_project ZILLURRAHMAN$ sudo ssh zillur@136.145.231.38
zillur@136.145.231.38's password:
Last login: Sat Dec 31 01:15:02 2016 from 67.224.154.165
-bash: /opt/sge/bin/lx-amd64/: Is a directory
[zillur@genomics ~]$ qsub
bash: qsub: command not found...
[zillur@genomics ~]$ systemctl start sgemaster
==== AUTHENTICATING FOR org.freedesktop.systemd1.manage-units ===
Authentication is required to manage system services or units.
Authenticating as: Zillur (zillur)
Password: 
==== AUTHENTICATION COMPLETE ===
Failed to start sgemaster.service: Unit not found.
[zillur@genomics ~]$ systemctl status sgemaster
Unit sgemaster.service could not be found.

Then again:

Code:

[zillur@genomics ~]$ echo "PATH=$PATH;/opt/sge/bin/lx-amd64/" >> /etc/environment
-bash: /etc/environment: Permission denied
[zillur@genomics ~]$ sudo echo "PATH=$PATH;/opt/sge/bin/lx-amd64/" >> /etc/environment
-bash: /etc/environment: Permission denied
[zillur@genomics ~]$ sudo echo "PATH=$PATH;/opt/sge/bin/lx-amd64/" >> /etc/environment
-bash: /etc/environment: Permission denied
[zillur@genomics ~]$ su
Password: 
[root@genomics zillur]# echo "PATH=$PATH;/opt/sge/bin/lx-amd64/" >> /etc/environment
[root@genomics zillur]# exit
exit
[zillur@genomics ~]$ logout 
Connection to 136.145.231.38 closed.
MacBook-Pro-55:final_project ZILLURRAHMAN$ sudo ssh zillur@136.145.231.38
zillur@136.145.231.38's password: 
Last login: Sat Dec 31 01:16:36 2016 from 67.224.154.165
-bash: /opt/sge/bin/lx-amd64/: Is a directory
[zillur@genomics ~]$ qsub
bash: qsub: command not found...
[zillur@genomics ~]$ systemctl start sgemaster
==== AUTHENTICATING FOR org.freedesktop.systemd1.manage-units ===
Authentication is required to manage system services or units.
Authenticating as: Zillur (zillur)
Password: 
==== AUTHENTICATION COMPLETE ===
Failed to start sgemaster.service: Unit not found.

What should I do now?

Best Regards
Zillur

MadeInGermany · 12-31-2016, 07:56 AM

There was a mistake in the advice given. Should have been

Code:

echo "PATH=$PATH:/opt/sge/bin/lx-amd64/"

Please change the semicolon to a colon! (with a text editor)

zillur · 12-31-2016, 12:56 PM

Thank you very much. I think bash_profile related problem is solved. Now I have another problem:

Code:

Unable to initialize environment because of error: Please set the environment variable SGE_ROOT.
Exiting.
99 jobs submitted as array task
Job ID:

Thank you again. What should I do now?

Best Regards
Zillur

szboardstretcher · 12-31-2016, 01:07 PM

You are going to have to read about the gridengine and how it works to decide where to put SGE_ROOT, i cannot do that for you. http://gridscheduler.sourceforge.net...e_shadowd.html

Setting it is as easy as the PATH variable, just throw it in bash_profile like:

SGE_ROOT=/some/path

MadeInGermany · 12-31-2016, 04:41 PM

I guess it is

Code:

export SGE_ROOT=/opt/sge

to be added to the bash_profile as well.

zillur · 01-04-2017, 02:33 PM

Thank you very much for your kind suggestions. after

Code:

export SGE_ROOT=/opt/sge

I got a new problem. Here it is:

Code:

[zillur@genomics idmapping]$ perl split_blast_queries_edwardsSteve.pl -f /home/zillur/Desktop/zillur/phd/orthofinder/fasta/PlasmoDB-28_Pfalciparum3D7_AnnotatedProteins.fasta -db /home/zillur/Desktop/zillur/phd/blast/blast/db/nrdb -d output/ -n 100
Unable to run job: denied: host "genomics.rrp.campus" is not submit host
Exiting.
99 jobs submitted as array task
Job ID:

Here is my host:

Code:

[zillur@genomics idmapping]$ hostnamectl status 
   Static hostname: genomics
         Icon name: computer-server
           Chassis: server
        Machine ID: ****
           Boot ID: &*****
  Operating System: CentOS Linux 7 (Core)
       CPE OS Name: cpe:/o:centos:centos:7
            Kernel: Linux 3.10.0-514.2.2.el7.x86_64
      Architecture: x86-64

What should I do now?
Best Regards
Zillur

John VV · 01-04-2017, 03:37 PM

you are aware that this is a CentOS 6 ( el6) rpm on CentOS 7 !

this will likely cause major problems
one being rhel6 uses systemV and rhel7 uses systemD

use the cent7 rpm ( el7 )

zillur · 01-04-2017, 03:58 PM

Thank you very much for your comments. How can I get el7 rpms? I got the latest rpm from here:

Code:

https://arc.liv.ac.uk/downloads/SGE/releases/8.1.9/

Best Regards
Zillur

John VV · 01-04-2017, 04:08 PM

grab the source and build it

rhel6 and rhel7 ( or centos and sl ) are VERY VERY different operating systems