In a world where application workloads can be in high demand at a moment’s notice, it is good practice to performance tune your Red Hat Enterprise Linux server infrastructure
An always-on Internet is enabling customers to interact with your organisation around the clock. To ensure the best possible experience, your server infrastructure should be kept optimised with regular performance tuning.
A number of emerging technologies and business processes require high levels of computer processing power, storage capacity and network bandwidth. These include:
IoT. The “Internet of Things” is a broad term used to describe connected devices and applications which can collectively send and receive massive volumes of data.
Big data analytics. With more data being generated at increasingly faster rates, big data analytics is being used to process raw data to deliver more real-time business insights.
Process automation. More process automation – including robotics – is calling for high-performance process control systems and data integration platforms.
Multimedia and VR. Audio, video and 3D imagery for augmented and virtual reality requires more grunt than standard text-based user interfaces.
Mobile apps and self-service. As applications go mobile and businesses begin to offer more customer and partner self-service, good performance is at the core of the user experience.
The self-service trend is particularly pertinent as customers and partners become more demanding of better application response times. Here, poor performance will have a direct impact on user experience and possibly revenue opportunities.
Factors like base system configuration, database tuning and application settings can impact how well the overall platform performs. And, if you are doing custom app development, the code itself should be optimised which adds another layer of complexity.
It is good practice to regularly performance tune servers and other software-defined services like networks and storage, as anything out-of-the-box tends to be geared for general purpose. Once you get into segmentation for specific workloads it makes sense to do performance tuning.
Getting performance tuning right will have immediate business benefits as you will be able to get the most out of modern server and storage hardware. The pace of hardware technology performance continues unabated and, when combined with optimised software for servers, storage and networks, will deliver continuous improvement over closed systems with limited flexibility or multi-tenanted public clouds.
Some tips for infrastructure optimisation
Let’s consider a number of practical steps to optimise server infrastructure.
Limit unnecessary services. If the server is being used for a dedicated workload then keep a close eye on redundant services which can impact performance.
Workload specific. Perform specific performance tuning for certain workloads. For example, database servers will have different configuration requirements to network servers. In the latest version of Red Hat Enterprise Linux you can optimise the system based on a profile which is already there. Previously, low-level work was required but now there is a wizard with some use cases around optimisation and RHEL performance tuning.
RAID and SSD. Use of solid-state disks has attracted much attention for its I/O and general performance benefits. This is true for most workloads; however, further performance benefits can be achieved by optimising the way all storage disks operate with RAID and modern file system technology.
High-speed interconnects. Like SSD technology, most network connections in the data centre can be sped up by using a technology with lower latency than the ubiquitous Ethernet standard. Most systems vendors are making high-speed interconnect technology available alongside Ethernet and it is a good time to take advantage of it.
File system maintenance. To keep a server or storage system running at its peak, the file system must be continuously monitored and tuned. Factors like file fragmentation and contention can degrade overall file system performance. A number of methods – from discarding unused blocks to changing parameters – can be used to prevent performance problems.
Two sets of eyes. When working on system performance tuning or code optimisation it’s always good practice to get a second opinion from a fresh set of eyes. This will provide a new perspective on your architecture and adds room for optimisation.
Provide feedback. If you are on a server support subscription then put it to good use and report your performance tuning work back to Red Hat. Contributing to the community process will also help improve the performance of future releases of the software.
As the world continues to run at a faster pace, performance tuning will make increasingly good business sense. Start by identifying the underperforming servers and consolidate them with updated equipment and server software.