To deliver your business-critical WordPress properties without fail Carbon60 focuses relentlessly on the four fundamentals behind our quality of service: latency, throughput, data scale, and reliability.
- Latency is the duration between the time that a user initiates a request, such as clicking a hyperlink, and the time until the last byte is transmitted to the Web browser.
- Throughput is the number of concurrent requests that can be processed without increasing latency beyond client defined acceptable thresholds.
- Data Scale is the amount of content that can be stored while maintaining latency and throughput targets.
- Reliability is the ability to meet your organization’s targets for latency, throughput and data scale consistently over time.
The ability of our Global WordPress Delivery Platform to meet your organization’s targets for latency, throughput, data scale, and reliability is a pillar of our industry-leading SLA.
Latency
Latency is composed of three major components which are especially relevant in scenarios where users are geographically distant from the source or have low-bandwidth network connections:
- The time it takes the system to receive user requests over the network;
- The time it takes the system to process and respond to user requests;
- The time it takes for the user’s browser to render the response.
Carbon60 engineered its WordPress Delivery Platform to minimize latency for local as well as geographically distant users by integrating a global content caching service which minimizes the latency by: 1) using a low-latency Tier 1 network for “first mile” Internet connectivity; 2) using best path networking protocols to minimize latency across the Internet’s “middle-mile”, and 3) directing users to content cached on their local networks. These techniques are proven to reduce latency by 10x and more.
Equally important for minimizing latency and achieving a great user experience is careful WordPress and site performance tuning. Our WordPress platform is carefully optimized so it responds and processes user requests as quickly and efficiently as possible. We also provide performance benchmarks and testing services so you can optimize your code and ensure you are meeting your first and second-access page load time targets for end users locally or globally at various load levels.
Throughput
Throughput is the maximum number of concurrent user requests your site can sustain without exceeding maximum latency targets. Every request has a specific cost in system resources. If there are choke points or the overall resource pool available is not large enough to sustain the number of concurrent request then new requests are queued. This cumulatively increases the latency of each subsequent request. If the queue grows too large then new requests will be dropped altogether. Before this point, however, the high levels of latency will have rendered the site unusable.
To maximize throughput so that your site can handle large traffic spikes seamlessly, Carbon60′s global WordPress delivery platform leverages an enterprise cloud computing infrastructure to dynamically allocate network, compute, and storage resources to your site as required. Moreover, the total throughput available on the platform and to your organization’s site is magnified exponentially by distributing it to tens of thousands of cache servers deployed globally. These caching servers reduce latency and increase throughput by offloading requests to servers close to the end user.
To handle massive volumes of concurrent requests on our WordPress platform your site’s code must be tuned to eliminate potential bottlenecks. Tuning your site includes: minimizing the “weight” of the most popular pages, minimizing round trip database calls, minimizing web service calls, optimizing database queries, loading data to the browser asynchronously, properly indexing database tables, and avoiding potential database locking issues. A common culprit of poor throughput is poorly written WordPress plug-ins and excessive “on-page” javascript. Carbon60′s performance testing tools and services are the best assurance that your code is up to the task.
Data Scale
Data scale refers to the amount of data that can be stored without impacting throughput and latency. Generally, the greater the volume of content – as measured by the number of database objects or sometimes files on a system – the greater the impact on overall throughput and user experience. The method that is used to distribute data across disks and database servers can also affect throughput.
To maximize data scale, Carbon60′s WordPress platform is built on an enterprise-class storage infrastructure that supports high I/O demands associated with large, busy databases. This infrastructure includes a 40 Gbps Fiber Channel over InfiniBand storage network attached to Tier 0 (SSD), Tier 1 (SAS) and Tier 3 (SATA) SAN storage. Each Tier of storage uses a distributed RAID 10 configuration to maximize read and write throughput across all available spindles. LUNS can be scaled to 100+ TB. Carbon60 continually optimizes application workloads and physical storage to ensure sufficient I/O reserves to manage peak loads.
Reliability
Reliability is the ability to meet established latency, throughput, and data capacity targets consistently over time. A reliable WordPress delivery service can consistently sustain latency and throughput targets during peak load and peak hours. It can also sustain latency and throughput targets during normal system operations such as routine maintenance, database indexing, site crawling by web bots and daily backups. To achieve “five nines” reliability, latency and throughput targets must also be maintained during extraordinary circumstances such as hardware failures, denial of service attacks and even natural disasters.
To achieve the highest possible levels of reliability under all circumstances Carbon60 uses many different technologies, tools, processes and technologies. For example, our capacity, change, and release management procedures maximizes reliability during normal system operations. And, to protect against extraordinary incidents, we have invested heavily in system redundancy and self-healing technologies as well as the ability to failover operations to secondary data center in the event of a disaster.
Quality of Service Guarantee
Carbon60′s focus on these four fundamentals – latency, throughput, data scale, and reliability – in tandem with our end-to-end technical support model is what defines our quality of service. It is why organizations trust us to manage their business-critical WordPress properties. This is more than just words, we stand behind our quality of service by including these fundamentals into our industry-leading Service Level Agreements.
Learn More
To learn more about Carbon60′s WordPress-as-a-Service please call one of our expert business consultants toll free 1-888-Carbon60 or email us at sales@carbon60.com
