Yes, we fit the first environment. We actually call it the "computer core" too, which is interesting. Our core consists of a Sr manager, two Windows server/desktop dudes, one VB programmer (for interfaces to instrumentation), and then myself (Sr Unix Mgr+sysadmin) and a Unix sysadmin (position currently open- hint hint hint - "inquire within"). Between our group, we manage all of the computing technology spread across several floors and a bunch in the data center. We do not use "corporate I.T." for anything except for network ports, because we're too specialized for them to be able to service the needs of the research community. They're uh.. a demanding bunch :-) Just as we treat the various sciences as specialties and have our own experts (genetic analysis, protein stuff, and all that - all the principle investigators to go fetch grant money and then do cool things with the computers...), we treat "computing" as its own specialty and have the in-house talent to manage it properly. I agree that when you end up with "research people" also doing double-duty as pseudo sysadmins, you end up with a lot of wasted time. --kcb -----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 9:30 AM To: Clustering, compute farming & distributed computing in life science informatics Subject: Re: [Bioclusters] Admins / node? Kent, You raise a really valid point. In my opinion there are two basic sorts of environment: 1) Your organization includes a research IT core. You're going to add a cluster to it. How many people do you add to that core? Or, if you're the core, how many new admins do you "bid" in order to agree that you can support this new system? 2) Everyone else. My feeling is that group 2 will require more person hours than group 1, since there are a lot of baseline networking / facilities / IT / account management responsibilities that come along with any big computing resource. -Chris Dwan On Feb 2, 2006, at 10:09 AM, Brodie, Kent wrote: > That's a really good question. I am also going to assume there's as > many answers as there are list subscribers. _______________________________________________ Bioclusters maillist - Bioclusters at bioinformatics.org https://bioinformatics.org/mailman/listinfo/bioclusters