Update: Jan/10/2012

• With the deployment of the new LDG-5.x, Condor no longer is installed via Pacman in the /opt/ldg directory.
• OS' native distribution [in the form of *.rpm packages for RedHat_based OS (Scientific Linux, CentOS, Fedora, etc) and/or *.deb packs for Debian] are used. All files are installed now in the Standard File Hierarchy [FHS] directories.
• Installation packs are downloaded from the Condor repositories, or
• For RedHat_based OS, a YUM repository exist at http://research.cs.wisc.edu/condor/yum/, which can be downloaded locally [by following the instructions at that URL] and then used to download Condor packages.
• For Debian, the APT repo exist at http://research.cs.wisc.edu/condor/debian/.
  
• There's no longer the need of sourcing "setup.sh" files to set any special environmental variable. System env vars will suffice.
• The following instructions will assume Condor version >= 7.6.x

---

Most likely the default installation and configuration is not what is needed.

As you must already know, a machine running Condor can play (basically) 3 roles:

• as the Central Manager [CM] node that rules the entire (condor) cluster,
• as a condor submission node which is used by people to send jobs to the (condor) cluster,
• and/or as a condor execution node which is the machine that executes the jobs sent by users.

Below are instructions for deploying Condor onto your cluster using one particular method and configuration option or style, and making some basic assumptions [a cluster will be configured with 1 CM node, 1 condor submission node, and several execution nodes]. Condor is very flexible and so you may choose to install, configure, and deploy it in a variety of ways. If you find that the instructions below do not suit your needs please see the latest Condor manual.

1. Install the rpm/deb condor package in ALL the machines of your cluster. The package will automatically add a 'condor' user/group if it does not exist already. Cluster sites with a specific security policy should add the 'condor' user/group manually before performing the installation.
2. Choose the machine that will play the role of the CM, and edit the main configuration file /etc/condor/condor_config:
   • Set CONDOR_HOST to the FQDN [Full Qualified Domain Name] of the machine that will play the role of CM. If this machine has a seperate network interface just for access to the cluster nodes, use that FQDN or equivalent IP address. For example, if that machine has FQDN universe.sverige.kth.edu and IP=666.999.555.111, then set CONDOR_HOST=universe.sverige.kth.edu or to its equivalent IP address, CONDOR_HOST=666.999.555.111.
   • Set RELEASE_DIR = /usr
   • Set LOCAL_DIR = /var
   • Set LOCAL_CONFIG_FILE = /etc/condor/condor_config.local
   • Set CONDOR_ADMIN to an appropriate email address
   • Set UID_DOMAIN to the subnet domain for your cluster. For example, at UWM a typical node has FQDN medusa-slave001.medusa.phys.uwm.edu so the subnet domain is medusa.phys.uwm.edu. This means, UID_DOMAIN=medusa.phys.uwm.edu
   • Set FILESYSTEM_DOMAIN = $(FULL_HOSTNAME) if your cluster does NOT have a shared filesystem for users, or set it to the subnet domain if it does have a shared filesystem for users.
   • Set USE_NFS = True if your cluster has a shared filesystem for users.
3. Create a tar file that you can deploy onto each node of your cluster, with the configuration file:

   tar -cf condor.tar /etc/condor/condor_config 
   

4. Deploy the tar file onto each node of your cluster, making sure that the deployed file are in the correct path [this means, condor_config will overwrite the old installed config file].
5. On each node execute chkconfig --add condor.
6. Back on the CM machine, edit the file /etc/condor/condor_config.local to add the following:

   COLLECTOR_NAME = $(CONDOR_HOST)
   DAEMON_LIST    = MASTER, COLLECTOR, NEGOTIATOR
   

   where the CONDOR_HOST value is taken from the same variable in the /etc/condor/condor_config file.
7. If this machine has a seperate network interface for the cluster nodes also add the line

   NETWORK_INTERFACE = your IP address
   

   which means the IP address for the seperate network interface.
8. Choose a machine to be the condor-submission node, and edit /etc/condor/condor_config.local to add the following:

   COLLECTOR_NAME = $(CONDOR_HOST)
   DAEMON_LIST    = MASTER, SCHEDD
   

   where the CONDOR_HOST value is taken from the same variable in the /etc/condor/condor_config file.
9. Choose the machines that will be the condor-execution nodes, and edit /etc/condor/condor_config.local to add the following:

   COLLECTOR_NAME = $(CONDOR_HOST)
   DAEMON_LIST    = MASTER, STARTD
   

   where the CONDOR_HOST value is taken from the same variable in the /etc/condor/condor_config file.
10. Start Condor first on the CM machine, with:

    /etc/init.d/condor start
    

11. Start condor on the condor submission node by doing the same.
12. Start condor on the condor execution nodes by doing the same on each node.
13. On the Condor Central Manager machine do /usr/bin/condor_status to see the status of your Condor pool. One must be able to see sth similar to

    Name               OpSys      Arch   State     Activity LoadAv Mem   ActvtyTime
    
    nemo-slave0001.nem LINUX      X86_64 Claimed   Busy     1.000  3960  0+06:42:47
    nemo-slave0003.nem LINUX      X86_64 Claimed   Busy     1.000  3960  0+09:14:30
    nemo-slave0004.nem LINUX      X86_64 Claimed   Busy     1.000  3960  0+06:41:30
    nemo-slave0005.nem LINUX      X86_64 Claimed   Busy     1.000  3960  0+03:47:21
    nemo-slave0006.nem LINUX      X86_64 Claimed   Busy     0.990  3960  0+00:48:46
    nemo-slave0007.nem LINUX      X86_64 Claimed   Busy     0.990  3960  0+00:36:10
    nemo-slave0008.nem LINUX      X86_64 Claimed   Busy     0.990  3960  0+00:31:38
    nemo-slave0009.nem LINUX      X86_64 Claimed   Busy     1.000  3960  0+02:47:22
    nemo-slave0010.nem LINUX      X86_64 Claimed   Busy     0.990  3960  0+05:15:34
    nemo-slave0011.nem LINUX      X86_64 Claimed   Busy     0.990  3960  0+06:41:03
    ...
    slot1_2@nemo-slave LINUX      X86_64 Claimed   Busy     1.000    25  1+01:02:19
    slot1_3@nemo-slave LINUX      X86_64 Claimed   Busy     1.000    25  1+01:00:46
    slot1_4@nemo-slave LINUX      X86_64 Claimed   Busy     1.000    25  1+00:59:34
    slot1_7@nemo-slave LINUX      X86_64 Claimed   Busy     1.000    25  1+00:56:14
    slot1_8@nemo-slave LINUX      X86_64 Claimed   Busy     1.000    25  0+10:49:23
    slot1_9@nemo-slave LINUX      X86_64 Claimed   Busy     1.000    25  0+10:47:48
    slot2@nemo-slave11 LINUX      X86_64 Unclaimed Idle     0.270  48397  2+03:25:40
                         Total Owner Claimed Unclaimed Matched Preempting Backfill
    
            X86_64/LINUX  5171     0    4931       234       0          6        0
    
                   Total  5171     0    4931       234       0          6        0
    

This completes a basic deployment and configuration of Condor. You are strongly encouraged to read the Condor Manual and learn how to configure Condor is the best way for your particular cluster.

Supported by the National Science Foundation. Any opinions, findings and conclusions or recomendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (NSF)

---

$Id$