Maximizing SQL Server on the HPE Superdome X

All database servers are not created equal. There are many commodity servers on the market, but when you need the maximum scalability, availability and flexibility possible, then you need to look for enterprise-level features that are not present in the low-to-middle-end commodity x86 servers. First, let’s take a quick look at some of the challenges that IT faces today when dealing with their mission critical databases and then let’s drill into some of the features that enable the HPE Superdome X and SQL Server to address these requirements.

Challenges for Enterprise-level Mission Critical Databases

Leverage Massive Scalability for In-Memory Technologies

The HP Superdome X CPU is based on an advanced x86 architecture that’s designed to support up to 288 cores and 24 TB of memory when using multiple partitions. This means that you can have up to two 12 TB Windows Server 2016 instances on one HPE Superdome X, providing extreme scalability and support for SQL Server’s in-memory technologies like In-Memory OLPT and the Columnstore index. Even so, CPU and memory performance would be restricted without extreme I/O capabilities to match. The HPE Superdome X utilizes four fully independent memory channels on each CPU socket, all of which run in parallel. HP Superdome X’s high performance I/O provides up to 800 GB of aggregate bidirectional I/O bandwidth to keep pace with its CPU and memory capabilities.

Provide Mission Critical Availability and Reducing Downtime

SQL Server’s AlwaysOn Availability Groups provide fast automatic failure as well as disaster recovery protection for multiple databases, but for true enterprise-level availability, the server must be as reliable as possible as well. The HPE Superdome X provides five nines (99.999%) availability for protection from unplanned downtime. Memory outages on the HPE Superdome X are reduced as a result of improved error handling capabilities that are supported by the Intel Xeon E7 V3 processors. In addition, Superdome X’s Firmware First architecture detects and contains errors before data can be corrupted and the Superdome X’s Error Analysis Engine (EAE) analyzes possible hardware errors and automatically initiates self-repair when possible, delivering true enterprise ready availability levels.

Increase Operational Efficiencies

The HPE Superdome X’s hard partitions (nPars) enable one HP Superdome X server to run several large instances of Microsoft SQL Server 2016 -- all completely independent of one another. Each nPar can be completely electrically isolated for one another. This level of in-box isolation enables you run OLTP and BI workloads independently out on the same HPE Superdome X. In other words, nPars allows you to cohost mixed workloads and divide them into isolated partitions. This enables you to streamline IT operations by reducing the number of required servers to support your diverse workloads, which in turn, reduces power and cooling requirements as well as decreases licensing costs. You can manage the different nPars using HPE’s Onboard Administrator (OA). The OA collects information from HPE’s Integrated Lights-Out (iLO) management processors. Companies can access the individual servers via iLO or see the entire HPE Superdome X system using the OA.

Sponsored by HPE & Microsoft