Built to spec; with plans to scale
Since coming on board with Rhubarb Media almost 3 years ago, I’ve worked with Chad and our team to produce great web sites for clients to market their services or even sell their products online. As newer web technologies, tools and frameworks have matured, many have become more comfortable with putting their money (and our time) into getting equipped with powerful and engaging web-based tools. Some like to use buzz words like “web 2.0″ or “cloud computing”, but it mostly boils down to features like web forms, embedding or integration of 3rd-party web services (such as weather from Environment Canada, SalesForce Web2Lead, Facebook & Twitter widgets, Snapengage, Disqus, etc…), database-driven content management systems and custom “widgets” or calculators to help their site visitors to better understand their business and open communion channels.
Sites with these types of rich features have a number of dependencies and layers of underlying systems to make them work. Like any system with dependancies (your car… the government…), they require maintenance, checks, balances and strategy to scale and accommodate new ideas and growth.
Growing pains
As usage of these systems increase, there are various areas in which you can “tune things up” to scale. Hardware and memory can be beefed up, caching can be applied at many many levels and code can be tweaked or even demolished to make way for new methods to increase efficiency. Further up the line new features could be requested or recommended to meet new needs. When it comes to the lower-levels of these systems, we like to recommend our own hosting services to better control the end-to-end experience and better serve our clients “when things go wrong”.
Yes, things can and do go wrong… and we own up when it’s on us. Increases in traffic and heavier usage of the various dynamic tools can contribute to systems hiccups and glitches. We see this all of the time on the web with services being unavailable due to many other moving parts (note Twitter’s famous Fail Whale; Chad also recently hit the wall on one of the e-blast services we rely on). Nothing is perfect and in the case of websites (with so many moving parts in both software and hardware) no web site can provide a guarantee for zero glitches and 100% uptime. Well, that sucks! But what we can (and do) strive for is excellence through learning from experience and in how we respond to issues when they come up.
We all need to take responsibility when times are tough. I know that Rhubarb assumes responsibility of managing the server our clients pay for in hosting and in responding to them with solutions to resolve an issue. In serving our clients we take the time to diagnose and identify quick-hits that we can simply take care of or apply a patch and move on, while other times it requires research, code… time. Ultimately, when an problem has been identified we want to build strategies and provide solutions to improve the situation long-term (patching is not a strategy, but it can “get the job done”).
“There is no free lunch, someone has to pay.”
Sometimes a developer’s response is “add more resources”; which is an action we are currently undertaking with server upgrades. While another response may be “hey, a new component could be put in place to automate ‘this thing’ since, in this case, we can clearly see a pattern.” Solutions like this take time (and budget). I’ve noticed that often times Rhubarb will go above the call of duty and absorb the cost to make things excellent.
When the time comes for us to stand back, identify or diagnose the problem and make recommendations to respond and improve, that’s where the technical (or even creative) bits end and good business relationships kick in. We love our clients and are here to serve them well!
As we work to scale our network infrastructure (strategy and details coming soon in another blog post), we invite our clients to follow @rhubarbmedia on Twitter.
– Tyler
