Monday Oct 13, 2008

Evolution of RAS in the Sun SPARC T5440 server

Reliability, Availability, and Serviceability (RAS) in the Sun SPARC Enterprise T5440 builds upon the solid foundations created for the Sun SPARC Enterprise T5140, T5240, and Sun Fire X4600 M2 servers. The large number of CPU cores available in the T5440 needs large amounts of I/O capability to balance the design. The physical design of the X4600 M2 servers was a natural candidate for the new design – modular CPU and memory cards along with plenty of slots for I/O expansion. We've also seen good field reliability from the X4600 M2 servers and their components. The T5440 is a excellent example of how leveraging the best parts of these other designs has resulted in a very reliable and serviceable system.

The trade-offs required for scaling from a single board design to a larger, multiple board design always impact reliability of the server. Additional connectors and other parts also contribute to increased failure rates, or lower reliability. On the other hand, the ability to replace a major component without replacing a whole motherboard increases serviceability – and lowers operating costs. The additional parts which enable the system to scale also have an impact on performance, as some of my colleagues have noted. When comparing systems on a single aspect of the RAS and performance spectrum, you can miss important design characteristics, or worse, misunderstand how the trade-offs impact the overall suitability of a system. To get a better insight on how to apply highly scalable systems to a complex task prefer to do a performability analysis.

The T5440 has almost exactly twice the performance capabilities of the T5220. If you have a workload which previously required four T5220s with a spare (for availability), then you should be able to host that workload on only two T5440s, and a spare. Using benchmarks for sizing is the best way to compare, and we can generally see that a T5440 is six times more capable than a Sun Fire V490 server. This will complete a comparable performance sizing.

On the RAS side, a single T5440 is more reliable than two T5220s, so there is a reliability gain. But for a performability analysis, that is contrasted with the fewer numbers of T5440. For example, if the workload requires 4 servers and we add a spare, then the system is considered performant when 4 of 5 servers are available. As we consolidate onto fewer servers, the model changes accordingly: for 2 servers and a spare, the system is performant when 2 of 3 servers are available. The reliability gain of using fewer servers can be readily seen in the number of yearly service calls expected. Fewer servers tends to mean fewer service calls. The math behind this can become complicated for large clusters and is arguably counter-intuitive at times. Fortuntately, our RAS modeling tools can handle very complicated systems relatively easily.

We build availability models for all of our systems and use the same service parameters to permit easy comparisons. For example, we would model all systems with 8 hour service response time. The models are then compared, thusly




Yearly Services

Sun SPARC Enterprise 5440 server

2 + 1



Sun SPARC Enterprise 5240 server

4 + 1



Sun SPARC Enterprise 5140 server

4 + 1



Sun Fire V490 server

12 + 1



In these results, you can see that T5440 clearly wins the number of units and yearly services. Both of these metrics impact total cost of ownership (TCO) as the complexity of an environment is generally attributed to the number of OS instances – fewer servers generally means fewer OS instances. Fewer service calls means fewer problems that require physical human interactions.

You can also see that the performability of the T5x40 systems are very similar. Any of these systems will be much better than a system of V490 servers.

More information on the RAS features these servers can be found in the white paper we wrote, Maximizing IT Service Uptime by Utilizing Dependable Sun SPARC Enterprise T5140, T5240, and T5440 Servers. Ok, I'll admit that someone else wrote the title...

Wednesday Aug 20, 2008

Dependability Benchmarking for Computer Systems

Over the past few years, a number of people have been working to develop benchmarks for dependability of computer systems. After all, why should the performance guys have all of the fun? We've collected a number of papers on the subject in a new book, Dependability Benchmarking for Computer Systems, available from the IEEE Computer Society Press and Wiley.

The table of contents includes:

  1. The Autonomic Computing Benchmark
  2. Analytical Reliability, Availability, and Serviceability Benchmarks
  3. System Recovery Benchmarks
  4. Dependability Benchmarking Using Environmental Test Tools
  5. Dependability Benchmark for OLTP Systems
  6. Dependability Benchmarking of Web Servers
  7. Dependability Benchmark of Automotive Engine Control Systems
  8. Toward Evaluating the Dependability of Anomaly Detectors
  9. Vajra: Evaluating Byzantine-Fault-Tolerant Distributed Systems
  10. User-Relevant Software Reliability Benchmarking
  11. Interface Robustness Testing: Experience and Lessons Learned from the Ballista Project
  12. Windows and Linux Robustness Benchmarks with Respect to Application Erroneous Behavior
  13. DeBERT: Dependability Benchmarking of Embedded Real-Time Off-the-Shelf Components for Space Applications
  14. Benchmarking the Impact of Faulty Drivers: Application to the Linux Kernel
  15. Benchmarking the Operating System against Faults Impacting Operating System Functions
  16. Neutron Soft Error Rate Characterization of Microprocessors

Wow, you can see that there has been a lot of work, by a lot of people to measure system dependability and improve system designs.

The work described in Chapter 2,  Analytical Reliability, Availability, and Serviceability Benchmarks, can be seen as we are beginning to publish these benchmark results in various product white papers:

Performance benchmarks have proven useful in driving innovation in the computer industry, and I think dependability benchmarks can do likewise. If you feel that these benchmarks are valuable, then please drop me a note, or better yet, ask your computer vendors for some benchmark results.

I'd like to thank all of the contributors to the book, the IEEE, and Wiley. Karama Kanoun and Lisa Spainhower worked tirelessly to get all of the works compiled (herding the cats) and interfaced with the publisher, great job! Ira Pramanick, Jim Mauro, William Bryson, and Dong Tang collaborated with me on Chapters 2 & 3, thanks team!

Tuesday Aug 05, 2008

More Enterprise-class SSDs Coming Soon

Sun has been talking more and more about enterprise-class solid-state disks (SSDs) lately. Even Jonathan blogged about it. Now we are starting to see some interesting articles hitting the press as various companies prepare to release interesting products for this market.

Today, CNET posted an interesting article by Brooke Crothers that offers some insight into how the consumer and enterprise class devices are diverging in their designs.  My favorite quote is, "One of the things that SSD manufacturers have been slow to learn (is that) you can't just take a compact flash controller, throw some NAND on there and call it an SSD," said Dean Klein, vice president of memory system development at Micron. Yes, absolutely correct.  Though Sun makes several products which offer compact flash (CF) for storage, the future of enterprise class SSDs is not re-badged CFs.  There are many more clever tricks that can be used to provide highly reliable, fast, and reasonably priced SSDs.

Monday Jul 14, 2008

Nice Article on Understanding Disk Reliability

David Stone of eWeek published a nice article this week on understanding disk reliability. It covers in a short, concise manner many of the aspects of disks that I've blogged about here.  Enjoy.

Examining Disk Reliability Specs

Wednesday Apr 09, 2008

RAS in the T5140 and T5240

Today, Sun introduced two new CMT servers, the Sun SPARC Enterprise T5140 and T5240 servers.

I'm really excited about this next stage of server development. Not only have we effectively doubled the performance capacity of the system, we did so without significantly decreasing the reliability. When we try to predict reliability of products which are being designed, we make those predictions based on previous generation systems. At Sun, we make these predictions at the component level. Over the years we have collected detailed failure rate data for a large variety of electronic components as used in the environments often found at our customer sites. We use these component failure rates to determine the failure rate of collections of components. For example, a motherboard may have more than 2,000 components: capacitors, resistors, integrated circuits, etc. The key to improving motherboard reliability is, quite simply, to reduce the number of components. There is some practical limit, though, because we could remove many of the capacitors, but that would compromise signal integrity and performance -- not a good trade-off. The big difference in the open source UltraSPARC T2 and UltraSPARC T2plus processors is the high level of integration onto the chip. They really are systems on a chip, which means that we need very few additional components to complete a server design. Fewer components means better reliability, a win-win situation. On average, the T5140 and T5240 only add about 12% more components over the T5120 and T5220 designs. But considering that you get two or four times as many disks, twice as many DIMM slots, and twice the computing power, this is a very reasonable trade-off.

Let's take a look at the system block diagram to see where all of the major components live.

You will notice that the two PCI-e switches are peers and not cascaded. This allows good flexibility and fault isolation. Compared to the cascaded switches in the T5120 and T5220 servers, this is a simpler design. Simple is good for RAS.

You will also notice that we use the same LSI1068E SAS/SATA controller with onboard RAID. The T5140 is limited to 4 disk bays, but the T5240 can accommodate 16 disk bays. This gives plenty of disk targets for implementing a number of different RAID schemes. I recommend at least some redundancy, dual parity if possible.

Some people have commented that the Neptune Ethernet chip, which provides dual-10Gb Ethernet or quad-1Gb Ethernet interfaces is a single point of failure. There is also one quad GbE PHY chip. The reason the Neptune is there to begin with is because when we implemented the coherency links in the UltraSPARC T2plus processor we had to sacrifice the builtin Neptune interface which is available in the UltraSPARC T2 processor. Moore's Law assures us that this is a somewhat temporary condition and soon we'll be able to cram even more transistors onto a chip. This is a case where high integration is apparent in the packaging. Even though all four GbE ports connect to a single package, the electronics inside the package are still isolated. In other words, we don't consider the PHY to be a single point of failure because the failure modes do not cross the isolation boundaries. Of course, if your Ethernet gets struck by lightning, there may be a lot of damage to the server, so there is always the possibility that a single event will create massive damage. But for the more common cabling problems, the system offers suitable isolation. If you are really paranoid about this, then you can purchase a PCI-e card version of the Neptune and put it in PCI-e slot 1, 2, or 3 to ensure that it uses the other PCI-e switch.

The ILOM service processor is the same as we use in most of our other small servers and has been a very reliable part of our systems. It is connected to the rest of the system through a FPGA which manages all of the service bus connections. This allows the service processor to be the serviceability interface for the entire server.

The server also uses ECC FB-DIMMs with Extended ECC, which is another common theme in Sun servers. We have recently been studying the affects of Solaris Fault Management Architecture and Extended ECC on systems in the field and I am happy to report that this combination provides much better system resiliency than possible through the individual features. In RAS, the whole can be much better than the sum of the parts.

For more information on the RAS features of the new T5140 and T5240 servers, see the white paper, Maximizing IT Service Uptime by Utilizing Dependable Sun SPARC Enterprise T5140 and T5240 Servers. The whitepaper has results of our RAS benchmarks as well as some performability calculations.




« April 2014