OpenVMS, Hewlett-Packard’s high-end operating system, got its first European outing this week on the company’s high-end Intel-based SuperDome server.
HP Brings OpenVMS to the SuperDome
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
These machines are nice and all, but they don’t hold up to clusters. They are great when you don’t need to scale beyond 64 procs though, if you can handle the price tag.
OpenVMS is where clustering was invented and perfected.
It is possible to cluster more than one of these beasts easily.
Is wonderful and all, but I think it will be ported JUST as poorly as 11I was to Itanium (We’ve put off our Itanium shift because it is so buggy that it is unusable.)
A 50K DL380 cluster would run RINGS around a $2M superdome.
http://tpc.org/tpcc/results/tpcc_price_perf_results.asp
We are currently looking to replace our N’s and RP’s with 380’s, and looking at a minimum of 10x the performance of our old solution. The older systems cost roughly $300K each w/ 8 550MHz chips, and my NOTEBOOK GROSSLY outperforms them.
im still skeptical on HP’s server plans. bring back openvms and tru64 on the alpha! >:(
Tim, I am guessing what you are REALLY saying is bring back the ALPHA, as both of the operating systems you mention are going to CONTINUE with the official HP roadmap
This is simply great news. If you don’t think that HP is committed to OpenVMS, I will ask you what other OS has been ported to three completely different cpus over the course of 20 years. The only other COMMERCIAL os that has this much porting was Winnt4.0, which ran on i386, alpha, powerpc and mips. However, that is no longer the case.
go HP!
I remember VMS to be a wonderfull OS.
Does is being activly developed? (not counting hardware support)
What the hell does that prove? It tells me that they can’t figure out where they want the OS to be, and they have managed to place it on 2 FAILED platforms. I have software that’s about 25 years old that still works on the latest rev of the same platform, how is this worse?
VMS is a beautiful OS and i am happy that it hangs on. And i find elements in Christopher’s post as well as in Anon’s from cable.mindspring.com to be supportive of both.
HP obviously recognized the value and power of VMS and sticks to it. OTOH, HP is apparently not sure were to go with VMS. Since it has been 64bit for quite some time (1995 i think) Itanium2 and Opteron sound logical.
HP is sure where to go with VMS. Its plan is to keep it alive for a long time. If you are referring to MPE, however , the other OS inside HP that has been EOL’ed
Again, OpenVMS will continue. What chip it runs on, as mentioned in the article, is pretty much a non-factor.
As for Mr Anonymous without the bravado to actually post using their name stating that there have been TWO failed platforms, do you consider Alpha or even VAX a failure? If either of these were even remotely failures (and they are NOT) then Openvms would have been dead at that moment. Ever think of it that way?
Alpha and VAX made Digital and then Compaq a large amount of revenue and HP recognized that and is keeping it going.
VMS is being slowly terminated by HP. HP is betting alot on Itanium and wants Alpha/VMS users to switch to Itanium based machines. Once this is done they will push HP-UX very hard on those customers and will completely kill Tru64 and VMS. VMS development is stagnant. They are simply porting VMS to other hardware to give those customers and upgrade option. They are strictly in “maintence” mode. The article even mentions DECnet, which is not something a new customer would really care about. HP is not targetting new customers to use VMS, just keeping it alive long enough to rope old users into switching to itanium, which will give them a good chance of moving HP-UX on to those machines.
Many of their customers are now switching from HPUX to Linux, the TCO numbers are so much lower that you can buy new x86 servers every year for the cost of maintenence on a single N-Class server. LOL
Silly HP! HP’s big VMS customers are there because a) the O/S is extremely reliable, and b) it does some things differently than Unix/Linux. If VMS goes away, there is no cogent reason to migrate the apps to HP/UX instead of another O/S, as the porting cost is the same.
Are you kidding me? HP is not interested in keeping OpenVMS in “maintenance mode”. HP MPE is in maintenance mode. No new features, and extended support for 5 years. OpenVMS is being ported to a completely new architecture. Do you have any idea of the engineering resources to do that? Its not a linux port to a new toaster folk. This is an enterprise OS as mentioned above, ready for mission critical applciations. The amount of TESTING of all the applications is enormous. The educations organizations that get new courses/certifications. Just think of the support organizations that will be trained on the new OS, all the work behind the scenes!
If you have facts, post them. Else, stop spreading illogical conclusions.
I don’t think that OpenVMS is best. After some time HP will decide to drop OpenVMS product!
Solaris is more better and I am recommending to migrate from HP Itanium systems to Sun SPARC systems!
People they are more more faster!!!
So, my anonymous, you are recommending sparc systems? You do know that Sun at some time till decide the drop the entire architecture?
At least HP has the enormous resources required to port to another CPU. Does Sun in its present financial situation? Its curtains soon for them. They certainly are not faster systems. Show me any SAPS scores that prove that please.
I worked at HP with engineers porting parts of VMS. Believe me they know their future. After the group that ported some of the libraries used on Itanium finished, most were reassigned to other projects. HP only has like 3 engineers around from a group of 20. Their sole purpose is maintenance/bug fixing once VMS on Itanium ships. HP is not comitted to adding new features, etc to VMS. Porting VMS does not require testing all applications that run on VMS. They are using many of the same tests used on Alpha builds that have been augmented with some itanium specific tests. Testing at HP (compaq projects at least) is a joke in general and the whole compaq comporate culture is rife with deadwood hanging on ’til retirement. HP’s merger has been a wakeup call to old DEC guys hanging around.
Once the port is done, HP will be reassigning a large portion of engineers to other projects and keeping around a few support engineers. HP is slowly killing all compaq related projects that don’t help their bottom line. The enterprise computing group is the ONLY part of HP to lose money during the quarter ending last fall. Carly herself said that the enterprise systems group will face some hard sacrifices, read: layoffs. HP doesn’t want to support/develop 3 or 4 enterprise OSes. They want to consolidate all high-end computing to HP-UX and everything else on the serverside to Linux and use Linux support revenues to help improve ESG’s bottomline.