Today is the Day…

Today is the day I decided to enter the 21st century, I finally decided to sign up to twitter. Obviously being a Sysadmin I’ve heard of this “twitter” milarky for a while but I could never bring myself to sign up. I tend to leave signing up to social networks quite late as typically they are fads, MySpace anyone?

Either way, Twitter has proven it is not going anywhere, not now and not in the immediate future, so I am taking this brave new step into the 21st Century; part of my brave new step is also this blog. For those that don’t know I did use to have a blog but that was back in the days when it was cool to have one where as these days it is more of a necessity, especially if you want to be taken seriously as an authority on a subject. With that in mind, this Blog is formed; my aim is write things that are not as personal but hopefully more useful than my previous blogging attempts.

Does Technology help us

So joining the 21st century as I have was to fix a niggling hole I felt I had in my online presence, I felt the urge to communicate the projects I’m working on and the day to day challenges that are faced in rapidly expanding environments.

As I said earlier I was late to the boat on Twitter, I had a blog, I stopped blogging and now have returned to blogging, I was late to Facebook, I was however very quick on MySpace.

This all got me thinking, so often as a sysadmin you are faced with technical decisions and the balance between stability and innovation.

Which leads nicely onto “How important are your clients?” and “What problem are you actually trying to solve?” a lot of the time a technology decision can be made in the heat of battle or with out considering all of the facts, Our jobs as sysadmins is to make that technology decision that is the best fit for the current problem, not one that may happen at some point.

So “Does Technology help us?” The short answer is no. Technology is not the answer to a problem, it never has been it never will be. Oh my, shock horro, A sysadmin who said technology is not the answer! Correct.

“Why is that?” you may be thinking, well most problems in IT are created by the technology that has come before hand. It is rarely caused by the Client or Customer asking for more features, and even when the request comes from Clients or Customers they often are not asking for a largely complex solution, lets examine this a little bit.

So your Client has asked for… “I’d like to have my website available in multiple countries”
For anyone that thought…”Oh host a server in which ever country, bingo” Bad Sysadmin! It is a website, it already is available in multiple countries. Our job is to work out what the client means when they ask these questions, depending on the answers the solution should be derrived. Did they just want a .de domain name pointing to the site and a German site presented? Did they want extra resilience?

Even if the client comes back with “I’d really like more resilience around my web server” we need to think about the why they would ask that, has the site been unstable lately? Are they expanding? Are they getting slow load times in the USA?

Our job is to ask questions and to provide the simplest solution to that problem. It may just be that we need to implement some monitoring as a first step or a better back up strategy.

Granted not every solution will be simple, you may need to set up some geo load balancing multi-site international data centre, or simply copy the site to another box in another country. Either way, start of simple, add complexity as needed.

Summary

I guess the point I’m getting too is don’t create a solution because you want to play with some new technology, don’t come up with a solution with out working out what the real point of the question was.

Adding in technology complicates a solution, so you should only add it in if it fixes a problem. In larger environments where you are supporting hundreds of servers you may need tools like Puppet, Chef, OpsView etc etc but in smaller environments you do not always need them.

The only exception to this is if you think there will be a need for it in the medium term, if you are rolling out a new solution that requires multiple environments and potentially will grow rapidly, then adding in these tools is a good start, but please start them out as simple as possible, you have all the time in the world to improve upon them later, don’t get caught up in technology for technologies sake, think simplicity, agility and rapid deployment.