Customers may like to know that we’re upgrading the Agilebase infrastructure, to generally speed up the system, make it more reliable and also ensure that as we grow, we can keep on top of monitoring and data security.

Technical Details

If you’re interested in these, read on. If not, feel free to skip to the bottom!

At the current count, we have 22 cloud servers, plus some additional services for monitoring, backups etc. They are of various types, including database servers, file storage servers, an AI server and the actual application servers running Agilebase (also including customer test servers).

The main improvement we’re making is that we’re now starting to manage all of these from a central ‘infrastructure as code’ system. It means that if we have many different instances of say, an application server, we can use one configuration to apply to them all. That means that when we make a change, e.g. to add a new security feature or firewall rule, it’s applied to all cases. There’s no danger of one being missed and potentially allowing a security or performance regression.

That also means we can with confidence build out our infrastructure to add more servers more easily, spreading the load between them and so improving performance.

One option that comes from this is that individual customers, if they want to, can have their own dedicated servers. As well as reliable performance, that gives them more control over restarting the system and upgrading Agilebase at a time which suits them. Please let me know if you’re interested in finding out more about that.

Finally the management system also lets us mix and match vendors. So one thing we’re doing is moving some services (less critical ones to start with) to a new hosting supplier Brightbox. They are a UK owned company and I’ve been very impressed with some of the innovative and forward looking things they’re doing in the hosting space.

That gives us the confidence of insulation from any future potential issues due to politics, e.g. tariffs or GDPR data sharing agreements being rescinded.

Actions

For the most part, we can carry out these upgrades with no disruption, simply starting up a new server, redirecting traffic from the old one and finally shutting the old one down.

From time to time though, we may contact individual customers to arrange some scheduling, most likely to put their system into a read-only mode for a short period to allow certain upgrades to complete. If so we will get in touch directly.

Get in touch

If you’ve any questions or just want to know more, please do contact us.

Many thanks to Ben Scattergood, who has been helping us with this process, introducing some cutting edge infrastructure management techniques!