[Bioclusters] Admins / node?

Brodie, Kent brodie at mcw.edu
Thu Feb 2 10:09:57 EST 2006

That's a really good question.   I am also going to assume there's as
many answers as there are list subscribers.

In our case:   I am the Sr. Unix sysadmin for our entire area (this
encompasses the dept of Physiology, but also the Human & Molecular
Genome Center, Bioinformatics Center, and a few other areas).    My
primary role and area of expertise, is the Unix sysadmin side.     My
job is to get the operating environment up and running for our
researchers, to KEEP it running, and to do all the standard sysadmin
"stuff": maintenance, patches, security, etc etc.   Under my belt are
dozens of servers-- several standalone systems, database servers, app
servers, misc servers, development environments, and we have Solaris,
AIX, and Linux to boot.   A little bit of everything.

In the case of our clusters, my role is getting it running-  that
translates to the operating system and tools (example: LAM-MPI in one
case).    I will also facilitate getting the applications installed
(compiled, built, distributed, etc).  Tools like "systemimager" make the
cluster deployment MUCH easier to manage.....

For me, that's where it ends.   I don't "use" the cluster, and in fact I
don't possess the skills or 'scientific knowledge' to run those tools
and to see if the results even make any sense.    We have really smart
users (you know, the types who all have multiple PhD's) who run the
tools, and let me know if everything is OK or not.   I monitor the
cluster's use, resources, uptime, etc.

As far as the number of people to manage the cluster environments- I
think that's wholly dependant on the complexity of the number of tools
and packages being run.    But, your math outlined in your original post
seems like a good rough estimate.

--Kent C. Brodie
  Medical College of Wisconsin

-----Original Message-----
From: bioclusters-bounces+brodie=mcw.edu at bioinformatics.org
[mailto:bioclusters-bounces+brodie=mcw.edu at bioinformatics.org] On Behalf
Of Christopher Dwan
Sent: Thursday, February 02, 2006 8:51 AM
To: compute informatics Clustering farming & distributed computing in
life science
Subject: [Bioclusters] Admins / node?

Question on the state of the art in cluster management:

Approximately what level of dedicated support do folks on this list  
have / wish you have for your clusters?  Obviously, there are a lot  
of free variables, including but not limited to:

* Does the support person also do development, parallelization, or  
otherwise *use* the cluster?

* Do their other responsibilities come from the IT side or the  
research side (i.e:  Are we dedicating half of a unix admin, or half  
of a postdoc?)

* How many users are being supported, and to what level?

Setting aside these and similar details that would make for valid,  
comparable numbers, my gut feel is that a reasonable guess is one  
half time IT person for the first fifty nodes or so.  I think it  
scales logarithmically from there:   So, go up to an entire full time  
person for 50 - 150, and add support staff incrementally as the  
cluster becomes more huge.

-Chris Dwan 
Bioclusters maillist  -  Bioclusters at bioinformatics.org

More information about the Bioclusters mailing list