Differences

This shows you the differences between two versions of the page.

--- progetti:cloud-areapd:user_s_guides:im_user_guide [2016/10/11 15:18] – [Troubeshooting] verlato@infn.it
+++ progetti:cloud-areapd:user_s_guides:im_user_guide [2017/07/21 08:03] (current) – verlato@infn.it
@@ Line 1: / Line 1: @@
+===== IM User Guide =====
+  * go to http://90.147.77.138/im-web/index.php and create an account <IMusername> <IMpassword>
+  * install in your machine the IM client (https://github.com/grycap/im-client)
+  * configure your im_client.cfg file as below:
+<code bash>
+[im_client]
+xmlrpc_url=http://90.147.77.138:8899
+auth_file=auth.dat
+</code>
+  * configure your auth.dat file as below:
+<code bash>
+type = InfrastructureManager; username = <IMusername>; password = <IMpassword>
+id = ost; type = OpenStack; host = cloud-areapd.pd.infn.it:5000; username = <yourCAPusername>; password = <yourCAPpassword>; tenant = GRID-CERTIFICATION; service_region = regionOne; auth_url = https://cloud-areapd.pd.infn.it:5000;
+</code>
+  * run the following command to create a CentOS7 torque/maui cluster with the master node with a floating IP by using [[http://pastebin.com/cp7G8rdN | os-c7-torque.radl file]]:
+<code bash>
+./im_client.py create os-c7-torque.radl
+</code>
+  * check on http://90.147.77.138/im-web/list.php the status of the infrastructure you just launched
+  * when the infrastructure is configured, login on the master:
+<code bash>
+$ ./im_client.py list
+Connected with: http://90.147.77.138:8899
+Infrastructure IDs:
+  c2f70fa0-8ba9-11e6-9dbf-0242ac110002
+$ ./im_client.py -a auth-os.dat sshvm c2f70fa0-8ba9-11e6-9dbf-0242ac110002 0
+Connected with: http://90.147.77.138:8899
+Warning: Permanently added '90.147.77.131' (RSA) to the list of known hosts.
+Last login: Thu Oct  6 09:53:03 2016 from 90.147.77.138
+[centos@torqueserver ~]$
+[centos@torqueserver ~]$ sudo su -
+[root@torqueserver ~]$ pbsnodes
+vnode-0-0
+     state = down
+     np = 2
+     ntype = cluster
+     mom_service_port = 15002
+     mom_manager_port = 15003
+vnode-1-0
+     state = free
+     np = 2
+     ntype = cluster
+     status = rectime=1475749622,varattr=,jobs=,state=free,netload=? 0,gres=,loadave=0.00,ncpus=0,physmem=4193904kb,availmem=3709676kb,totmem=4193904kb,idletime=2323,nusers=0,nsessions=0,uname=Linux vnode-1.localdomain 3.10.0-123.9.3.el7.x86_64 #1 SMP Thu Nov 6 15:06:03 UTC 2014 x86_64,opsys=linux
+     mom_service_port = 15002
+     mom_manager_port = 15003
+vnode-2-0
+     state = free
+     np = 2
+     ntype = cluster
+     status = rectime=1475749628,varattr=,jobs=,state=free,netload=? 0,gres=,loadave=0.00,ncpus=0,physmem=4193904kb,availmem=3714272kb,totmem=4193904kb,idletime=2319,nusers=0,nsessions=0,uname=Linux vnode-2.localdomain 3.10.0-123.9.3.el7.x86_64 #1 SMP Thu Nov 6 15:06:03 UTC 2014 x86_64,opsys=linux
+     mom_service_port = 15002
+     mom_manager_port = 15003
+[root@torqueserver ~]# su - dteam001
+[dteam001@torqueserver ~]$ ls
+test.job
+[dteam001@torqueserver ~]$ qsub -l nodes=2 test.job
+.torqueserver.localdomain
+[dteam001@torqueserver ~]$ ls
+errori  risultato  test.job
+[dteam001@torqueserver ~]$ cat risultato
+Thu Oct  6 10:25:55 UTC 2016
+Ci sono una serie di cose interessanti che ti interessera sapere
+Questo job stato identificato come 0.torqueserver.localdomain e si chiama ExampleJob
+e stato inserito inizialmente nella coda batch
+ed e stato eseguito sulla coda batch
+E stato sottoposto dalla macchina: torqueserver.localdomain
+E stato eseguito sulla macchina: vnode-2.localdomain
+Thu Oct  6 10:26:00 UTC 2016
+</code>
+  * the master will take one of the 3 available floating IPs: 90.147.77.131-->133, corresponding to cld-cream-01-->03(.pd.infn.it) in the DNS
+  * X509 host certificates are available for cld-cream-01-->03 hosts
+  * an example of .radl file for creating an HTCondor cluster on CentOS7 is here: [[http://pastebin.com/qiT9Vszf | os-c7-htcondor.radl]]
+  * sometimes the cluster do not configure properly. Use e.g. $ ./im_client.py destroy c2f70fa0-8ba9-11e6-9dbf-0242ac110002 to destroy the cluster, and try again.
+  * after destroying a cluster, do not forget to re-allocate the floating IP in the GRID-CERTIFICATION project: login as root in cld-ctrl-01 and execute the script ./GRID-CERTIFICATION-floIPs.sh
+===== Troubeshooting =====
+  * to login the VM hosting IM server execute the following from cld-ctrl-01:
+<code bash>
+[root@cld-ctrl-01 ~]# cd MARCO
+[root@cld-ctrl-01 MARCO]# ip netns exec qdhcp-b6e3d4de-2959-4c89-8558-5845f2b316fd ssh -i for_im_web_key  centos@10.63.34.5
+or
+[root@cld-ctrl-01 MARCO]# ssh  -i for_im_web_key centos@90.147.77.138
+[centos@im-server-and-web ~]$
+</code>
+  * to restart the service:
+<code bash>
+[centos@im-server-and-web ~]$ sudo su -
+[root@im-server-and-web ~]# docker stop im
+[root@im-server-and-web ~]# docker stop im-web
+[root@im-server-and-web ~]# docker start im
+[root@im-server-and-web ~]# docker start im-web
+</code>
+  * to login in the container running the service:
+<code bash>
+[root@im-server-and-web ~]# docker exec -ti im /bin/bash
+root@e78904a1fcc6:/#
+</code>
+  * in case you need to reset everything:
+<code bash>
+[root@im-server-and-web ~]# docker stop im; docker rm im
+[root@im-server-and-web ~]# docker stop im-web; docker rm im-web
+[root@im-server-and-web ~]# docker run -d -p 8899:8899 --name im grycap/im
+[root@im-server-and-web ~]# docker run -d -p 80:80 --name im-web --link im:im grycap/im-web
+</code>