Stratus ships latest batch of fault-tolerant Xeon servers
Fault-tolerant computing veteran Stratus has released the latest generation of its ftServer systems, which offer zero downtime for mission-critical applications, but lag behind the rest of the market in terms of the latest technology.
Stratus has announced availability of the 12th generation of its ftServer platform, claiming the systems offer up to 25 percent greater performance over the previous generation thanks to more CPU cores and expanded networking connectivity.
There’s just one snag – instead of running the newest 4th Gen Xeon Scalable processors, which Intel introduced in January, the updated ftServers appear to be based on 2nd Gen Xeon Scalable processors, now several years old.
But it should noted that these Stratus systems are based on an unusual fully redundant architecture which doubles up the processors, memory and storage, and ensuring that new hardware meets the mission-critical reliability requirements will be tougher than simply turning out a standard x86 box.
Wordle recreated in Pascal for the Multics operating system
In practice, this architecture means that each ftServer consists of two identical customer replaceable units (CRUs), which are essentially separate Intel-based server nodes that run in lockstep so that if one fails, the other continues running.
The 12th generation ftServers comprise four models; the 6920, 6910, 4920 and 2920, configured with 10 up to 44 cores in each CRU. The 2920 is a single-socket design that targets remote offices and branch offices, while the remainder are all dual-socket systems, with the 4920 targeting medium-sized facilities.
The 6920 is the high-end system configuration intended for data-intensive and transaction-intensive workloads, including payments processing, but also AI and ML applications, Stratus said.
This can be configured with up to 1280GB of DDR4 memory and up to 8 SAS drives per CRU, with up to 4 U.2 format NVMe SSDs an option. Each CRU also has two gigabit Ethernet ports and two 10Gbit ports, plus a 10/100 management port. Full specs can be found on the Stratus website.
Although not the latest technology, Stratus touts these systems as the company’s most powerful ftServer systems yet. “With this release, we are supporting the current and future compute needs of our customers,” said Stratus veep of strategy and business line management Jason Andersen.
It adds support for VMware vSphere 7.0, for example, although vSphere 8.0 is now available.
- Intel pulls plug on server system design division
- Arm, Intel make it easier to churn out Arm SoCs from Intel fabs
- Intel ships multi-die chips ahead of schedule – to the US military
- Intel successfully ships an updated datacenter roadmap
Andersen claimed that the new boxes offer more versatility and performance for the growing demands of a variety of applications, from transaction and payment processing to industrial automation.
Stratus states that its redundant architecture means there is no interruption in processing, or loss of performance or data integrity if a component fails. Unlike a high-availability cluster, an ftServer will continue to function while any issue is being resolved, it claims.
The company cites 99.999 percent uptime or better for its systems, or less than five minutes of downtime annually on average. Stratus said its platforms are backed by customer services that include 24x7x365 monitoring and management services via a secure, private network, while additional services offer access to “availability experts” for custom application development and infrastructure design consultation.
Virtualization drives demand for ftServers
However, while this kind of redundant system was vital in the past for organizations running mission-critical workloads, there is now competition from datacenter platforms such as VMware, which offers high availability support via its ability to migrate virtual machines to a new host server in the event of a hardware failure, for example.
Omdia chief analyst Roy Illsley agreed, telling us that modern demand for this type of server is limited.
However, he added that with the rise of edge computing, such redundant systems could become more relevant, especially in situations where there may be a single server installed in a remote or difficult to access location.
“I think it will be relevant for those deployments where only a single server is installed and the connectivity is poor – meaning protection by replication to another site is not reliable. Beyond that I think it is a specialist market,” he said.
These ftServer models are unlikely to be cheap as well. The company has not quoted any prices for the new systems, but previous generations cost in the range of tens of thousands of dollars. ®