Desktop Virus Protection, Blade Virtualization & More Reorg!
Jun. 14, 2006 by ravishan
More Reorg
Steve submitted his reorg plan which he has already discussed with his staff. This will go into effect officially on July 1. As per this plan:
- Darrell will report to Annette
- Doug and Sharon will report to Jane
- Rich and Gaylord will report to Dan
- Chip wil lreport to Barbara
In many of these instances, the projects that everyone was working on already had this relationship established, though not in any official capacity. Steve also is workign with HR on some title changes to reflect this. We will let you know once this is firmed up.
Desktop Virus Protection
We had a couple of very interestign presentations to attend this week.
We have been using a virus scanning/removal software from a company called Sophos for the past several years on all our mail servers successfully. It works just the same way the virus removal software works on the desktop. Virus definitions are downloaded to our servers on a regular basis and we use a public domain software called Mailscanner, which in concert with Sophos software scans each mail message coming in for any viruses or any suspicious file names and removes the infected attachments.
The same company has a desktop version that runs on Windows machines and Macs (yes, they are anticipating that some of the holes in the MacOS will soon be exploited). It goes beyond the traditional virus protection and also does spyware and adware protection. They demonstrated their central management console and Joanne and James seem to like it. We are in the process of negotiating a price… Since vendor negotiations is in my portfolio, I will be pressing them hard.
Blade Virtualization
Another demo that we saw was from a company called Egenera. Steve Windsor and Tom Dimauro recommended that we look at them as an option for Oracle 10.g. The technology that this company has developed is exciting. I will call it Blade virtualization to keep it simple. What they have is a set of diskless blades and one or more master blades all connected to SAN (storage area networks). Conceptually, think about their technology as one where the masters are configured to distribute the applications amongst the blades in such a way that there is a high level of redundancy and load balancing. An application at various times may be running on different blades depending on load, for example. As a user no one will know any of these and their experience will be unchanged. By making them diskless, you are achieving even more ease of administration by having only one active copy of the OS in the SAN which they use to boot. This system is also energy efficient.
But there are a few problems… the cost for one. For 6 blades the approximate cost is $80K. We pay considerably less for 6 blades. There is no easy transition path from where we are to this. In other words, we cannot take what we have today and convert them to be Egenera compliant. So, we have to start over from here and commit to it going forward.
I do not believe that the actual return on investments they are talking about applies to us. If you are interested, I will be happy to talk to you about it. It is also my opinion, that there will be others (possibly public domain) who would be offering the concept of ease of management and virtualization for much less or part of the Linux OS in the next few years because these are real issues….

In regards to virtualization and management, it might well be worth our while to investigate Xen. While not providing the full feature set of something like our dear friends at Egenera, it gives virtual servers (through software) with little performance hit (compared to something like vmware). Similar idea (in fact, Egenera mentioned they were partnering with Xen at some point, not sure if you were there then..) but free and I think I saw gui management tools.
V mentioned it specifically as a tool used heavily within environments he works in, and investigations show that it seems to be quite popular for having multiple “servers” on one blade.
If nothing else, I suspect it will be adopted (if it hasn’t already) by more mainstream Linux resources in the near future (if it hasn’t already) so it’s worth watching for a special RedHat edition or what have ye if not letting us loose upon the technology ourselves.