The recent ruling on the Oracle vs. HP case throws a lifeline to those customers already committed to Oracle on Itanium, like my current company, but what does this really mean as far as new customers are concerned?
Now this is just my opinion and you are free to disagree, but as far as I’m concerned, making the decision to go Oracle on Itanium is a massive mistake. In fact, it has been a mistake for the last 5+ years. During the death of Tru64, HP sales people were advising us to go HP-UX on Itanium and HP engineers were advising us to go Linux on x86-64, at approximately 1/10 of the cost. Who are you gonna believe? I’m sure there are workloads where Itanium used to work well, but from what I can see in my company, the regular Linux on x86-64 kit kicks the ass off the HP-UX on Itanium kit. When you consider the price difference between the two, that’s really the final straw…
I’ve now lived through the death of Oracle on Tru64 and I’m experiencing the slow death of Oracle on HP-UX. Forcing Oracle to support Itanium is all well and good, but it doesn’t make it a good choice for the future. Like Tru64, new versions and patches of Oracle products take ages to appear and the level of support for these platforms has always lagged far behind. I’ve heard plenty of people make the same complaints about Oracle on AIX, but I’ve got no recent experience of this, so it’s just hearsay.
When I’m asked my opinion, this is pretty much what I say:
- For any 3rd party product (that includes Oracle DB), always pick the platform the stuff is developed for. In this case, I would suggest that Oracle Database on OL/RHEL x86-64 is the natural choice. Pretty much everything else is a port from there. Solaris (SPARC) has some life, having been bought by Oracle, but that would still be further down the list for me, especially as the hardware is crazy expensive. Solaris on x86-64 is such a small user base I don’t even count it yet.
- The Oracle middleware products are Linux-only gear IMHO. Yes, they are supported on other platforms, but personally, I would not go near them. I’ve had a couple of bad experiences here and it’s just not worth it. Once again, new versions and patches come out quickest on OL/RHEL, so it’s a no-brainer for me.
To emphasize the first point, we recently initiated a project for a non-Oracle 3rd party product and were told it was supported on Linux or Windows. I asked the question, “What do most of your user base run on?” The answer that came back was Linux, so we went Linux. If they had said Windows, we would have gone Windows. Of course I have my preferences, but I don’t want to be anyones guinea pig where production applications are concerned!
When I arrived at my current company, the general strategy was Oracle DB on HP-UX and middleware products on RHEL. We will have some HP-UX kit for a few years to come, but Oracle databases for all new products are going on to Linux from now on. Most of those will probably be on VMware, since that is a strategic platform for our x86-64 installations. It looks pretty likely we will ditch RHEL in favour of Oracle Linux. Suits me just fine!
Note. The legalities of what Oracle did is not my concern in this post. The courts say they were in the wrong and they will have to pay for what they have done. My point is from the perspective of a user of Oracle products on this platform.