Skip to main content

Last week we turned off our first ever dedicated server, Spindogs1. It had served us reliably over nearly a 10 year period but after completing a successful migration to a new, more powerful, replacement it was finally time to pull the plug.

Spindogs Server

Why now?

A lot has changed since we originally purchased Spindogs1, as we now actively maintain 24 servers for our clients. In 2015 we also added Steve Tozer, our system administrator, to the team because the demands and responsibilities of maintaining these servers had grown into a fulltime requirement.

When we first turned on Spindogs1 there was just five of us, and now there are over 20 and we have a support team, led by Angharad Yates, that in 2015 alone fielded 2585 client requests (that’s approximately one every working hour of every day!) with an average close time of under five days.

What’s new?

Our servers are located in a secure datacentre in Gloucester, as the bandwidth required to serve all of our websites would far outstretch even the powerful 40mb dedicated line we have in our office. Although our servers have always had a Linux operating system, over the year our preferred distribution has become CentOS, which is an enterprise class platform that ensures it has longer life cycles for updates and maintenance updates, which is crucial for ensuring that our servers remain secure in the future.

For increased redundancy and data protection all of our servers now have at least RAID 1 as standard, which means that all data on the server is automatically replicated across two hard drives. This means that if one hard drive fails the server will continue to operate as normal.

Our backup procedures have also improved over the years. Our backups are held in a different location to the servers, are monitored daily to ensure they are running and we carry out bimonthly scheduled disaster recovery routines, whereby we test recovering websites from a backup to ensure that the backups are working as intended. We keep the last four nightly backups on a rolling cycle and these currently take up over 2TB worth of disk space and contain 26878061 files – a long way on from our first website launch which was in total under 10mb in size!

How does it affect performance?

All of our servers are also monitored for uptime as standard and we are immediately notified by email and SMS if there is ever any disruption to the service. This helps us achieve our 99% uptime comfortably and, in fact, in 2015 all of our servers were online for over 99.9% of the time.

So, as we say goodbye to Spindogs 1, it will be interesting to look back in another 10 years and see how much further things have moved on from where we are now…

How can we help?