|
Tandem Computers was an early manufacturer of fault tolerant computer systems, marketed to the growing number of transaction processing customers who used them for ATMs, banks, stock exchanges and other similar needs. Tandem systems used a number of redundant processors and storage devices to provide high-speed "failover" in the case of a hardware failure, an architecture that they called NonStop. Over the two decades from the 1970s into the mid-90s, Tandem systems evolved from custom hardware to commodity CPU designs. The company was eventually purchased by Compaq in 1997 in order to provide that company with more robust server offerings. Today it is still known as NonStop, as a separate product line offered by Hewlett-Packard.
History Tandem Computers was founded in 1974 by a group of engineers from Hewlett-Packard, led by James Treybig. Their business plan called for systems that were safe from "single point failures" that were only slightly more expensive than competing non-fault tolerant systems. Tandem considered this to be very important to their business model. Limiting the additional expense was important since customers often developed procedural solutions to failures when the price of fault tolerant hardware was too high. The first system was the T/16 (later renamed NonStop I). The system design was complete in 1975, and the first example was sold to Citibank in 1976. The NonStop consisted of between 2 and 16 processors, each capable of about 0.7 MIPS with their own memory, I/O buses, and dual connections to their custom inter-CPU computer bus, Dynabus. The modules were constructed with dual paths so that any single failure would always leave at least one bus (both I/O and Dynabus), free for use by the other modules. The CPU was patterned on the HP3000 CPU, a microprogrammed 16-bit stack-based machine with 16-bit user addressing. Like the HP3000, the NonStop CPU added a number of registers for fast access, such as base addresses for global and local variables. The NonStop I ran a custom operating system called Guardian. It supported a NonStop programming paradigm that allowed a program to be completely fault tolerant. Several other companies introduced failover technologies but only Guardian supported completely fail-safe transaction processing. A properly constructed Guardian program could restart from any point and resume transaction processing without any loss of data. While conventional systems of the era, including mainframes, had failure rates on the order of a few days, the NonStop system was designed to fail 100 times less, with "uptimes" measured in years. Nevertheless the NonStop was deliberately designed to be price-competitive with conventional systems, with a simple 2-CPU system priced at just over two times that of a competing single-processor mainframe, as opposed to four or more times of most competing solutions. NonStop I was followed by the NonStop II in 1981, a slight improvement in speed to 0.8 MIPS, but a more measurable upgrade in memory from a maximum of 384 kB per CPU in the I, to 2 MB in the II, and the addition of a complete virtual memory system allowing for considerably larger address spaces. The same basic system, including the physical packaging, was used in 1983's NonStop TXP system that over doubled the speed to 2.0 MIPS, and the physical memory to 8 MB. In all of these machines the same Dynabus system was used, which had been overdesigned in the NonStop I so they could avoid changing it in the future. Introduced along with the TXP was a new fibre optic bus system, FOX. FOX allowed a number of TXP and NonStop II systems to be connected together to form a larger system with up to 14 nodes. Like the CPU modules within the computers, Guardian could failover entire task sets to other machines in the network. The company attempted to grab a piece of the rapidly-growing personal computer market in 1985 with its introduction of the MS-DOS based Dynamite PC/workstation. Sadly, numerous design compromises (include a unique 8086-based hardware platform incompatible with expansion cards of the day and extremely limited compatibility with IBM-based PC's) relegated the Dynamite to serving primarily as a smart terminal. It was quietly withdrawn from the market within a short period of time. In 1986 a major upgrade to the system was introduced, the NonStop VLX. VLX used a new Dynabus, increasing speed from 13 Mbit/s to 40 Mbit/s (total, 20 Mbit/s per independent bus). They also introduced FOX II, increasing the size of the networks from 1 km to 4 km. Using the original FOX VLX systems could be used with the older NonStop II and TPX's, but these systems were not supported on FOX II. VLX was partnered with the NonStop CLX, a minicomputer sized machine for smaller installations. The CLX had roughly the same performance as the earlier TXP, but was much smaller and less expensive. By the end of its lifetime the CLX had increased in speed considerably, and competed with the VLX, 1991's CLX 800 was only about 20% slower than the VLX, with the main difference being more limited expansion abilities. In 1986 Tandem also introduced the first fault-tolerant SQL database, NonStop SQL. Developed totally in-house, NonStop SQL included a number of features based on Guardian to ensure data validity across nodes. NonStop SQL was famous for scaling linearly in performance with the number of nodes added to the system, whereas most databases of the era had performance that plateaued quite quickly, often after two CPUs. A later version released in 1989 added transactions that could be spread over nodes, a feature that remained unique for some time. Later, the SQL database group was first co-opted then absorbed into Microsoft's SQL development effort. One outcome of this collaboration was Microsoft's clustered system technology. The NonStop Cyclone was introduced in 1989, introducing a new superscalar CPU design. It was otherwise similar to earlier systems, although much faster. In general terms the Cyclone was about four times as fast as the CLX 800, which Tandem used as their benchmark. On the downside the new CPU was complex and expensive, requiring four circuit boards to implement a single CPU. In 1991 Tandem followed this with RISC-implementations of Guardian, running on MIPS R3000-based CPU modules in the Cyclone/R and CLX/R. Programs written for the earlier stack-based CPU design were automatically translated on the fly into R3000 code in an interpreter, although they ran considerably slower than on earlier machines. Tandem also provided a number of tools to easily port existing object code to the new systems, resulting in code that was some 25% slower than the original Cyclone. Source code compilers were also available. While slower, the new system was considerably less expensive, and it was clear that RISC performance was outpacing CISC. By making the move when they did, they were banking on increases in MIPS performance quickly wiping out any performance disadvantages the system had at the time. In 1993 the NonStop Himalaya K-Series using the MIPS R4400 was shipped. In 1997 Tandem introduced the NonStop Himalaya S-Series. The S-Series machines were the first systems that changed the underlying architecture of the NonStop family, basing both the I/O and inter-CPU communication on their new ServerNet interconnect. Whereas Dynabus and FOX linked the CPU's together into a ring, ServerNet was a true point-to-point network replacing both, and ran at much higher speeds. ServerNet later was used as the basis of the InfiniBand industry standard. The S-Series machines continued the use of MIPS processors, including the R4400 and R10000. All the more recent systems were based on microprocessors, and the internal circuits of these chips are not fully checked. To assure correct computation, each logical processor had two microprocessors operating in lockstep. If the results coming out ever disagreed, the processor was considered to be faulting and instantly stopped. At that point Guardian would move that task to another processor as in earlier systems, guaranteeing that bad data was never written out due to hardware failures. A different approach was used in a separate family of computers, the Integrity line. These computers used additional redundant CPUs running the same instruction stream. When a fault was detected (e.g. by lockstep mismatch), the failing module was disabled but the redundant module continued processing the instruction stream without interruption. Since this was handled primarily in hardware, it could be used with a slightly modified conventional operating system; Integrity used a Unix variant rather than Guardian. The line was introduced in 1990, apparently as a response to the machines of Stratus Technologies (which were remarketed by IBM as System/88). Although distinct from the NonStop line, the Integrity designs were also based on the MIPS processors. With the introduction of the Integrity S4000 in 1995, the line was the first to use ServerNet and moved towards sharing hardware designs with the NonStop line. Tandem was acquired by Compaq in 1997. Compaq was in turn acquired by HP in 2002, bringing Tandem back to its original roots. As of 2003, the NonStop product line continues to be produced, under the HP name. After being acquired by HP, the NonStop line has moved to Itanium based processors, called Integrity NonStop Servers. The original Integrity line is no longer produced but the name 'Integrity' has been adopted by HP for all Itanium based servers. The modern NonStop platform incorporates a POSIX compatible environment (OSS) and Java. There is also an effort by HP to run Linux on the NonStop hardware. Also, Linux or other Unix based operating systems could be installed on the NonStop platform via a virtual machine environment. | ||||||||
|
| |||||||||
 |
|
| |