In Part 1 of this piece, we essentially talked about why the speed of high performance infrastructure is important, tools to quickly test your site, and how a faster site specifically boosts revenue. Now let’s discuss steps you can take – beyond infrastructure – to accelerate your site, followed by another reason you need strong, reliable hardware: business continuity.
- Beyond infrastructure, how do you get fast?
- High performance infrastructure: key to business continuity
- HA as fundamental to high performance
- Vow to be redundant
- Is your load imbalanced?
- Um, did we mention CDNs or the cloud?
Beyond infrastructure, how do you get fast?
Performance must be considered from multiple angles. Along with internally implementing or working with a web host that has high performance infrastructure, here are a few additional steps you can take to get your site moving, highlighted by Sherice Jacob on the Kissmetrics blog.
- Tell your site to gzip it. Many web thought-leaders recommend compressing responses using this common method. “Compression reduces response times by reducing the size of the HTTP response,” notes Chris Coyier of CSS-Tricks. “Gzip is the most popular and effective compression method currently available and generally reduces the response size by about 70%.”
- Crunch your images. You can slim your images with the “Save for Web” feature in Fireworks and PhotoShop. If you are graphically underendowed, you can go to it.
- Don’t expect HTML to do the heavy lifting. HTML allows you to adjust size once you have something uploaded (as through the WordPress UI). Bear in mind that the browser still loads the image at full size prior to resizing it, though.
- Cache yourself. WordPress and other CMS platforms have caching plugins that set aside the latest version of your site so that the page doesn’t have to populate from scratch with every browser request. WP Super Cache is widely used.
- Beware complex detours. You want to retain SEO you’ve built and modify your structure with 301 redirects. However, a jumble of redirects results in latency.
Beyond all the DIY steps you can take above to improve your speed, Jacob also mentions one infrastructural component that she believes is critical: a content delivery network (CDN). “Content Delivery Networks work by serving pages depending on where the user is located,” says Sherice. “Faster access to a server near their geographical area means they get the site to load sooner.”
High performance infrastructure: key to business continuity
It’s boring to think about infrastructure for most people, because it sounds like a bunch of machines and wires that are shut away in some warehouse, and that are simply linear conduits for the thought of the animate bipeds. Think about this, though: it is that infrastructure that allows your business to function and operate on a moment-to-moment basis!
After all, digital reality doesn’t just connect you with customers through content and e-commerce but with your colleagues. Consider how reliant you are on email and project management or other collaborative software.
Since we have become so dependent on these tools in an effort to increase efficiency, the high availability (HA) that is inherent in high performance infrastructure becomes a central concern.
HA as fundamental to high performance
HA isn’t optional but necessary if you want to maintain business continuity in a well-integrated, connected company. In other words, you need your infrastructure to suffer very little downtime.
A sound high-availability strategy “detects points of failure that can potentially cause the downtime and mitigates failure by distributing the load and traffic across the infrastructure,” notes TechAcute. “In the event of failure, a high availability infrastructure will have failover and recovery mechanisms.”
There are numerous reasons why you might experience downtime, because of failures in different parts of your system, such as:
- Operations; or
- Internal programs.
Your downtime could be because of interaction with your website from customers as well. You might have a spike and go down if your server isn’t prepared for Black Friday, for instance. After all, 30% of retail sales each year occur between Black Friday and Christmas, according to Kimberly Amadeo in The Balance. Let’s put that into perspective: If that period of time were even with the rest of the year, it would represent about 8% of sales. The actual 30% results are 3.75 times beyond the 8% expectations, so that’s a 275% increase in expected sales for the average e-commerce site. In the Black Friday economy, 100 lava lamps translate into 375 lava lamps. In an infrastructure that isn’t high-performance and readily scalable, you can’t keep up with that pace.
Your site might also become unreliable because you get hit with a DDoS attack or experience other hacking activity. In other words, security is an element that must be built into a high performance infrastructure.
Why is avoiding downtime so important, whether it’s caused by a flood of real or phantom traffic? “Aside from loss of potential sales, customers might not trust your brand or business in the future,” explains TechAcute. “Similarly, a business using an enterprise platform to manage its resources will compromise the integrity of internal communications.”
One other aspect of the availability that you achieve with a high performance infrastructure is that you are able to meet the expectations of the service-level agreements you hand to your customers. That’s just one more reason you never want your infrastructure to be the weakest link.
Vow to be redundant
If you’re writing an English paper, it’s fair of the professor to dock you for being redundant – because in that context, repetition isn’t appreciated. However, you want repetition, i.e. redundancies, and failovers in your infrastructure so that the system has alternatives when parts malfunction.
“Redundancy is having extra components available in the case a component fails,” notes Brian Heder in Network World. “Failover is the mechanism, be it automatic or manual, for bringing up a contingent operational plan.”
These two elements must be considered in the interest of the HA you achieve with a high performance infrastructure.
Is your load imbalanced?
When you talk about using a system that has strong load balancing, that means that you are using the simple tactic of distribution to your advantage. By distributing systems, you build a huge amount of redundancy into the system, but you really want to make sure your load is balanced across all your hardware.
“Cheap datacenter hosting will not accommodate a surge of users or other factors that can put a heavy strain on the servers,” says TechAcute. “Overload in the servers will cause an online service to go down.”
Load balancing means that your traffic is evenly running through various servers, so that your system is naturally more available and can maintain great speed.
Um, did we mention CDNs or the cloud?
As indicated by Jacob above, a CDN can be powerful in delivering high availability: it limits the distance between any user by utilizing datacenters in broad geographical locations, improving how quickly a page loads case-by-case.
To optimize high availability, it’s best to pair a CDN with the cloud. “Cloud platforms are perhaps the most cost-efficient solution in bringing about high availability,” advises TechAcute, “because your business does not have to invest in the capital expenditure required to purchase, run and maintain hardware.”
High performance infrastructure can increase your productivity and revenue, as well as maintain the trust and credibility of your brand. At Total Server Solutions, our high performance SSD-based cloud and CDN grow with you. See how.