Boston Linux & Unix (BLU) Home | Calendar | Mail Lists | List Archives | Desktop SIG | Hardware Hacking SIG
Wiki | Flickr | PicasaWeb | Video | Maps & Directions | Installfests | Keysignings
Linux Cafe | Meeting Notes | Blog | Linux Links | Bling | About BLU

BLU Discuss list archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Discuss] Education - becoming a Cluster Administrator- Type of Cluster is HPC

Hello All,
    The type of cluster is a HPC.
    Jobs are submitted to the nodes through SGE from the head to one of 12 nodes.
    Hardware question - answered by a suggestion of using Amazon or Access to Vmware.

    I have access to Vmware as well as a Dell Workstation with 12G of Ram and about 1 TB space.

    Going with the Vmware idea it sounds like a better option is to create a virtual head node and maybe three nodes all in 
Vmware. Include a means for centralized logins for this environment as a start.

    Wish to make it a Debian environment.

    Continue to research.

  ----- Original Message ----- 
  From: Stephen Goldman 
  To: discuss at 
  Sent: Thursday, December 22, 2011 10:49 AM
  Subject: Education - becoming a Cluster Administrator

  Holiday Greeting to all!
      My manager suggested I learn how to become a "linux cluster administrator". The question at hand is how best to do it?
      My organization currently has access to a cluster environment that is supported by an outside consultant.

       I have an account on the system which consist of an Ubuntu head node and 8 other servers under it. I manage shared storage that attach to the cluster but have no admin access to it.
      Essentially, I am a team of one and train myself.  I am not a Linux wizard. I support structured cabling, Windows desktops, Windows servers, Linux storage servers running RHEL 5.5 -6.1. Continuing to improve my skills as needed to fit the environment. Additionally I maintain a back end network, small machine room and Cisco 2800 router.

      The direction is to build a small test cluster.There is room in the machine to add test equipment and access to ip address. 
  The operating system of the cluster will be Ubuntu 10.04 - and to install SGE to manage jobs. The question is obtaining the hardware .. maybe from Ebay. I looks like I would pay for it or maybe find another more cost effective way to train. That is the issue. If it come down to it I am willing to invest in my training.

  Please share your thoughts..



BLU is a member of BostonUserGroups
BLU is a member of BostonUserGroups
We also thank MIT for the use of their facilities.

Valid HTML 4.01! Valid CSS!

Boston Linux & Unix /