Something cluster-like this way comes

It appears that Microsoft is about ready to release CCS. This might be interesting, depending upon what was done, and how it all works. Some things to note. It includes a job scheduler. Built into the OS. This is either a really good thing, or a really bad thing. I can see arguments both ways. It includes bits that Linux has had for a while. Remotable/scriptable installation. Multiple security models.
Some thoughts. The Platform folks in Toronto cannot be happy about the integrated scheduler. Microsoft is firing across the bow of them, Altair, and others. We will have to see how interested the customer base is in this.
Can you imagine MCSEs looking after your cluster? No, really. Now stop laughing.
My questions: how are they going to deal with virus scanners, firewalls, and the myriad of other requirements that windows machines must have in all corporate environments? How are they going to compete costwise with Linux? If they are not giving it away for free, then the likelyhood of it catching on in various submarkets (the fast growing ones) is about nil. No, I don’t have to pay RedHat, SuSE, or anyone else to build a cluster today. If someone wants to make my customers pay for the honor of doing so, then …
… they need to have a really compelling case in terms of functionality, support, features, etc.
This is where the CCS argument has not made sense to me in the past. You have an entrenched zero acquisition cost, low support cost, high functionality, high extensibility, highly open, large and (insert your favorite adjective for intense preference) group of followers … with more software being ported daily, more code developed for it … Again, what is the argument for CCS?
This is the point I was trying to make last year at SC05. I will make it again now. I am not convinced that there is a business case here for them. I think there are ways that they can play together better, but I am quite skeptical of this amounting to much of anything. I could be wrong, though I am hardpressed to see how.
Other contenders have come and gone, and some are still here but fading. Sun pushes Solaris 10 everywhere. I grabbed it and installed it. My impression is not alltogether favorable. I didn’t get a sense of “better than linux”, rather quite the opposite. My best experience with Solaris was the OpenSolaris Benelinx bootable CD. That worked (Solaris took about a week to install on hardware that linux took to without a problem, and the Benelinx CD booted to quite nicely). Speaking with customers and end users, I simply don’t see much call for Solaris in HPC. Yes there will be a few. Sort of like the few windows clusters. There is a reason there will be a few, and it doesn’t have much to do with quality. It has to do with marketing.
Will CCS be a hit? I don’t know, I think it is an uphill battle at best against an entrenched enemy which is growing and covering more ground against it. Solaris isn’t IMO relevant in this market (let me know if you think otherwise, I would like to hear some different perspectives). It is going to require some really clever marketing and pricing to have this make sense to users.

3 thoughts on “Something cluster-like this way comes”

  1. Certainly. Would be great to have real information, and not pure speculation. Would be better still to play with it. We tried Solaris 10, we tried many different Linux distributions, we would be happy to try WCC.
    Contact data is landman _at_ scalability _dot_ org .
    Better yet, feel free to compose a short post as to why you believe the product is beneficial, why it is needed, what differentiates it from what is out there.

  2. Joe – tomorrow (June 9) Windows CCS will be released to manufacturing, meaning that the code is done and headed out the door to OEMs. We expect general availability (volume licensing) in August, with OEM solutions from Tyan, Verari, Dell, HP, IBM and others shortly thereafter. Sorry, no retail version. You and your readers can go to for the eval software, white papers, community site, etc. No benchmarks yet, but we’re hopeful of cracking the Top500 later this month. Most benchmarks will come later this summer and into the Fall.
    So why in Windows CCS beneficial, and why is it needed? Here’s what a few early adopter customers have said:
    “we???re able to leverage of our existing Microsoft infrastructure and skills to reduce costs and improve security via identity management.”
    “because Windows has broad support from application vendors, we???re able to partner with BioTeam to quickly build and deploy the drug discovery application.???
    ???In addition to serving massively-parallel applications, I???ve found that Windows Compute Cluster Server is a convenient tool for serving the computational needs of many small projects, where installing the software, updating databases and managing other such tasks are much easier than on a set of separate computers.???
    “we ported WRF to Windows CCS because I continually receive requests for Windows-based WRF from end users at research labs, commercial weather forecasting companies and academic institutions who have Windows-based workstations.”
    Those comments come from people at the Genome Research Institute at University of Cincinnati, Cornell University???s Computational Biology Service Unit, and National Center for Atmospheric Research.
    I’ll point out a few other places that you and your readers can monitor and participate in order to assess the need for Windows CCS. First is a blog site for 10 supercomputing academic centers running Windows CCS.
    See here: Not all of the bloggers there are proficient as Joe 😉
    I’ll also point you toward the Windows CCS newsgroup:
    And a community site just kicked off last month:
    I’ll check back next week, and I’d be happy to provide more specifics from you or your readers.

Comments are closed.