A number of years ago, I was working booth duty at Linux World in San Francisco. I believe that I was showing off the HP Virtual Server Environment for Integrity servers running Linux, now Insight-Dynamics - VSE for Integrity servers
At one point during the day, after a customer demo, one of the two people watching the demo started asking about HP-UX 11i, and UNIX in general. We talked about server virtualization for HP-UX for a bit, and then came the big question. Does UNIX have a future? Isn't Linux on x86 going to wipe it out?
My answer then, much as it is now, was that UNIX is a large, mature market, and that the existing vendors were going to keep innovating in that market for years to come. The revenues on non-x86 servers, at least back then, were roughly the same as for x86 servers, although the x86 server market was and is growing faster. Linux is going to grow, especially for new workloads. However, UNIX will still deliver mission critical environments, especially for existing workloads, better than many of the operating systems typically found on x86 servers. I used server virtualization as a proof point: the UNIX environments were offering a variety of mission critical virtualization technologies that were not available on Linux at that time (such as dynamic hard partitions and instant capacity), and indeed, are still not available today. In other words, Linux was small and growing, but UNIX was big and stable, and would be around for years to come.
After the customer walked away, satisfied with the answer, the other person watching the demo spoke up. She worked on AIX, and commented that she was glad I got to answer that question, since she also had to answer it many times! We shared a laugh, and went our separate ways.
Now, years later, I still get the question on a regular basis, and my answer is still much the same. I keep speaking with sales reps, who are still making a living selling UNIX. I keep speaking with customers, who definitely want more UNIX servers, and are often wondering not whether or not to keep UNIX, but which UNIX vendor they should buy from (HP of course, since that is where I draw a paycheck <grin>).
In addition, we have the Gabriel Consulting Group reports on UNIX preference . Their report on UNIX is title UNIX: Alive, Well, and Strategic . In short, the report states something that many of us already know: In the datacenter, UNIX is still a go to choice for mission critical workloads, especially for large enterprise customers.
I look forward to working with those customers and sales reps for years to come.
Recently, we were running to power and performance benchmarks on a server in our lab. We were using HP-UX 11i v3, update 5 and we were testing the impact of some of the power management features on the system.
The power management features that are available as of the HP-UX 11i v3 Update 4 release (0903 release) include Green Idle processors (pwr_idle_ctl), Green Active Processors (pstatectl), and more (see http://h71028.www7.hp.com/enterprise/w1/en/os/hpux11i-power-cooling-overview.html).
Green Idle processors basically takes the idle processors and drops it into a low power idle state, also known as the c-1 state. This is the same idle state that unutilized Instant Capacity processors are placed into automatically by the firmware. One interesting note is that when the OS places today's Itanium Montvale processor into C-1 state, it wakes up the processor every 10 milliseconds to handle a timing interrupt (if I remember correctly... I know the 10 ms is correct, but it could be a different interrupt). Anyway, basically, it wakes up the processor every 10 ms, at which point, it checks to see if there is any queue for it process, and if not, drops back into a lower power idle mode.
Green Active processors takes working processors and drops them into lower power p-states, basically slowing the frequency of the active processors. While the p-state can by set manually (to high performance or low power modes), the most balanced mode is the dynamic mode. This basically changes the p-State based on the activity level of the processor. If utilization is low, it drops it into a lower power mode. If it is in the 50% range, it increases the frequency, and around 80% utilization, it puts the processor in high power mode (P0 state).
So, why the explanation? Well, we were testing a system, and with Green Active processors and Green Idle processors enabled, it's maximum performance was within ~1.2% of our base run (and since we only ran it once for each run, it could be within the run to run variation). But here is the big difference: with both Green Active and Green Idle processors enabled, power consumption at 100% utilization dropped by about 8.5% and power consumption at idle dropped by about 11.7%
Now, I know that people who purchase mission critical systems want to get the maximum performance from those systems. However, would they be willing to activate these power controls since it has a minimal impact on performance? Even more importantly: we ship systems in maximum performance mode today, and those people who install the HP-UX 11i v3 updates that have these features can rest assured that the default is the same high performance mode that their previous releases that didn't have these commands used.
My question: have you enabled these power management features on your systems today? If not, why not? Should we change our maximum performance defaults, and ship our systems with power management enabled, and turn on dynamic power management with HP-UX 11i v3 future updates? Comments and feedback would be greatly appreciated.
Oh, and if you are interested in learning more? Check out the web site listed above, or the white paper at http://h20195.www2.hp.com/V2/GetDocument.aspx?docname=4AA2-5482ENW&cc=us&lc=en.
We finished a fabulous Superdome Tech Day yesterday with a number of bloggers. I need to go and see what they have written, but I know that at least one site actually covered the event live. There were also a lot of tweets using the tag #HPSuperdome.
There were a couple of questions, thoughts, and interesting things that struck me.
- OpenVMS is on HP Integrity servers. A few of the bloggers thought that OpenVMS was dead, and they were glad to hear that it is still alive, well, and being developed.
- HP Non-Stop servers are also Integrity servers, using the HP Itanium processor. They aren't Superdomes, but this was also something that a lot of people we not aware of in the room.
- I had a great chat with Jean Bozman from IDC after her presentation. It was interesting to note that the UNIX server business has consistantly been 31-32% of the overall server market revenue since they started counting the market in 1996 up until 2008. The overall market goes up and down, and the UNIX business with it. However, this explains to many people why UNIX servers will be around for a long time. The Windows and Linux on x86 markets have definitely grown, but it was mostly at the expense of "Other Servers" while the UNIX business has remained steady.
- We had a chance to look at a Superdome. Not just a new Superdome, but an original pre-production Superdome delivered months before the product was announced. It doesn't have any of the skins on it, but has been upgraded over the years with all the new components. It's still used for performance work today, almost 10 years after it was first installed.
- We finished the day with a quiz on the benefits of moving from an IBM mainframe to an HP Superdome. The questions were based on facts that come up during things like the Mainframe TCO challenge . We had a lot of fun with that, but some of the TCO savings by moving to a HP Superdome just blew people away.
Are there any of these topics that you want to hear more about? Leave me a comment, and I'll see if I can address them in more detail in a future blog post.
After taking a few days off and heading home to Canada for a wedding this past weekend, I'm back in Calfornia and back in the office. Today is a busy day. We're hosting a number of blogs and technical publications for a Superdome 10th Anniversary kick off.
Superdome came out in the fall of 2000. In fact, I remember standing on a tradeshow floor (in Philidelphia) watching the live kickoff of this new server from New York. 9 years later, we're kicking off the Superdome 10th Anniversary celebrations by inviting a number of blogs for a day of discussion around Superdome. They include OS News , TechVirtuoso , Unix Adminosphere , SDR News , Active Win, and Cuddletech .
We're talking about such trivia as where the Superdome name came from (project name was Half Dome... but some of our collegues didn't want to sell half of anything), the history of the Superdome from PA-RISC to today's Itanium processors), HP-UX 11i and server virtualization. Superdome has definitely come a long way... more than 20x the performance, running a few hardware partitions to up to 2500+ virtual machines today, and up to 2 TB of memory today. We've heard some interesting customer stories, seen some neat pictures (I'll probably share a few of them later if I can), and seen some demos.
As we look forward to celebrating the Superdome 10th anniversary, it brings up a few questions. Do you use a Superdome? If so, what do you use it for? What is your most interesting Superdome story?
I hope to share a few as we go along.
Did you know that there is an Energy Star certification for servers? More information is at http://www.energystar.gov/index.cfm?c=ent_servers.enterprise_servers. Did you know that HP had the first Energy Star certified servers (several HP ProLiant servers)?
I attended a stakeholders meeting with the EPA last week to discuss their proposals for Energy Star Tier 2 . It was a productive meeting, and I was there to try and get more servers, such as blade servers and servers with more than 4 sockets to be covered by Energy Star.
However, this brought up an interesting question, particularly for mission critical servers. While some of the requirements are for manufacturing issues, such as efficient power supplies, there are also requirements for power reporting, and likely with Tier 2, active power management. This raises an interesting question. We have recently introduced some power management features in HP-UX 11i v3 (updates 3 and 4, http://h20341.www2.hp.com/enterprise/w1/en/os/hpux11i-power-cooling-overview.html). Some features, such as Green Idle Processors and Green Active Processors (also known as c-states and p-states) can reduce power consumption today and have minimal to no performance impact (ex. <10 ms), except in situations where very fast response time is key. By default, we've disabled the power management settings, since they might potentially impact performance. The EPA will likely require them to be turned on by default if we want to use them for qualification purposes.
Should HP just enable dynamic power management, with little to no performance impact, for all servers? Or should we ship all servers set to have the maximum performance by default, as we do now? What do you think?