New office success!

Posted on December 29, 2010

A fairly unique opportunity arose for me towards the end of 2010: I was able to design, purchase components for, and then oversee the implementation of a new office infrastructure. This was due to a new physical build, and based on the success of the previous SAN and Infrastrucure project, I was essentially given full control over all aspects of the IT infrastructure.

It was an extensive piece of work and covered about six months of my life before finally going live at the end of November 2010, it involved visits to the site before building works were completed; setup of an entire dedicated test lab; and various correspondances thrasing out the business requirements for the new build.

The actual move-and-go-live happened over a weekend, the staff downed tools on Friday, had their PCs moved over the weekend and were able to log on and start work again Monday morning, a huge achievement bearing in mind that the entire system had changed: everything from the cables to the printers to the back end storage was now completely different.

Two things made this project enjoyable and successful in my eyes:

1. Specifications for the IT infrastructure were reviewed with open minds rather than penny-saving, “what’s-wrong-with-what-we’ve-got” ones. This is partly due to the success of the previous infrastructure project, but also partly due to the fact that I was able to show, was able to prove, that the solution I was suggesting was the right one for the job.
2. Site visits to thrash out requirements and check progress. Every single trip to site, even when I arrived to find the people I was due to meet weren’t coming or something equally as time-wasting, proved to have been worth it, every single trip allowed opportunities to correct things that hadn’t gone quite right, or ask the right person the right question to make sure things were perfect on time.

Just a quick list of some of the individual technical successes in the office that I’d like to share with you and perhaps serve as a checklist for getting a new build / refresh right:

• Server Room: Obvious to some, seemingly unnecessary for others. If you don’t have a server room of some description then you’ll find it harder to enforce environmental control (temp/hum) and physical security is lowered (may not be an issue if you trust all your staff, but might be an issue if the building gets broken into). Additionally, the more servers you have, the noiser it’ll all become, no-one wants to sit next to that.
• Environment: Firstly, get some way of controlling the environment in terms of dust, temperature and humidity. Remember that “keeping servers cool” isn’t really what air con is for, what it’s actually for is “keeping the air at a constant, suitable temperature and humidity”. An air con system that kicks in at 30 degrees C and then lowers the room to 5 degrees C will break your servers faster than none at all. In my case I had to push the point of air conditioning as the building designers were convinced that passive extraction would be suitable. I was able to prove my point by simulating the heat output in a room of similar dimensions with no cooling: the temperatures rose to an unhealthy level in just over 30 minutes.
• UPS: get an uninterruptable power supply for the servers, the size and duration of backup will depend on your sector and the service agreements you want to honour, but at the very least allow the servers to shutdown gracefully. In this case I comissioned a UPS which included a maintenance bypass switch (if UPS completely breaks and needs to be swapped out this can be done without downtime now) and a temperature sensor which will initiate a shutdown if the server room gets too hot (i.e. if the cooling breaks). As this is a virtual environment I’ve tied the UPS shutdown signal to VMWare rather than the windows machines.
• Wireless/WLANs: Not high on many people’s agendas, but it’s surprising how ineffective “normal”, household access points and routers are for medium-sized businesses. Quite often if you connect 5 or so wireless clients then the access points simply can’t cope and the connection speeds drop right down. This office doesn’t use wireless as the primary LAN connection but we wanted to make sure it was future proofed, and make sure that the load of visiting clients didn’t affect the private network. Due to the timing the WLAN also runs at 802.11n and n draft 2 (up to 600mb/s).
• The Servers: This is a bit of a “catch all” for everything else, but in short the solution put in place offers 1GB to each desktop with no bottlenecks; the servers are fully redundant, as is the data storage; off-site backups are taken automatically to an office over 50 miles away for disaster recovery and on-site backups are taken with snapshots several times a day. There are single points of failure in this office, such as the WAN link, but as the data is stored locally on site (due to size) the office would still be able to complete a large percentage of work while the link was repaired (and in the last year it hasn’t gone down once, so the risk there is acceptable).

Not one of those things above came without a small amount or discussion, but in each case it was relatively easy to show that the money spent upfront would be very much worth it over the next five years.