It's exceeded our expectations. After Oracle brought its 'unbreakable Linux' campaign to LinuxWorld in the summer of 2002, a lot of our customers were convinced that Linux was for real. In 2003, bigger customers started deploying Linux in production. In some cases, they had road maps of moving every single server they had to Linux; in others, it was moving new systems. Now, pretty much every customer that I talk to has Linux as one of their strategic platforms. Have there been any surprises for you in terms of Linux adoption?
The surprise is that people have really saved a lot of money on their infrastructure. We knew that it was cheaper than buying the big boxes, and you can run clustering stuff on cheap hardware, get more scalable components and pay as you go. Since the hardware is so cheap, there are more deployments happening much sooner than we anticipated because management doesn't object to an investment that is like twice the cost of a desktop. Older and smaller projects that were left behind because they were so expensive to get started now are coming up. I've heard that Unix is losing the most ground to Linux. Have you seen it that way?
The CPUs are really fast, so they can do way more with a small system than the equivalent with a traditional Unix environment. But there are a lot of people switching from the Windows environment, so it is not easy to say that Unix is the big loser. I don't think that is the case. From what we are seeing, the growth is everywhere. That is another good sign -- that we are not necessarily destroying one thing or one vendor, that it is an across the board migration.
Some customers still want the big SMP systems. They still want this 32-way machine, and Linux hasn't run on that. Linux 2.6 will be released in the first production version in the next month with SuSE, and it will be able to go to the high end. So, I think the next step for the biggest companies is to use the 32-way machines with Linux. A big step for our market is going after the high-end server.
Office suites are still maturing, and I think they still need some time. Engineering desktops, however, are happening today. For the rest of 2004, what new technologies should IT pros check out?
Generally, there are more Linux applications for them to evaluate, as ISVs are moving their applications. I think that a lot of other customers have worked with them to convince them that they should move. Of course, the big ISVs are there, but even the smaller ones have realized that you know they really can make money on Linux; it is a fast growing platform they should be there.
Virtualization again is something that is being worked on but it is kind of hard to tell when it will happen. It's a cool technology, and there's lots of interest, but it is not very clear when it will go strong.
But I have to keep going back to the importance of 2.6. SuSE's Linux 2.6 distribution is going to be pretty cool to see. 2.6 has all of the stuff that people need. It is very mature and stable and has all of the functionality needed in the enterprise. We were almost there in the past, but we are there technology-wise now. I think from that now people can calm down because this is totally equivalent to the other operating systems out there. They should no longer worry that everything that you need is there. What leading edge technologies are going to come on strong?
The big one is NUMA support. Linux 2.6 has support for NUMA, so you will even see this on 64-bit machines. Now that 2.6 has support for that, Oracle can make use of that and make better decisions of where to start up on the machine and optimize it better. So, it is a very much needed technology for kernel functionality on high-end systems, and it is there now.
FEEDBACK: What Linux-related technology is your company planning to evaluate or deploy in the next six months?
Send your feedback to the SearchEnterpriseLinux.com news team.