Since 2012 our main infrastructure has been based in Ireland, specifically in the AWS Ireland Region. This has allowed us to scale, with demand as well as comply with various laws and policies which help keep your data safe. For a start, your data never leaves the EU, this means that we fully comply with the Data Protection Act and ensures that your data never hits the USA, where data security laws are quite different to ones within the UK/EU.
So, why have we moved to London?
Being based within Ireland has worked well, however ideally we would have liked to have been based within the UK, where the majority of our customers are based. On the 13th December 2016, Amazon announced a new London Region. We knew that this was coming within 2016 so have already been planning for a move to this new region, once it was launched. As well as moving to London, we would also use this opportunity to upgrade our whole infrastructure, streamline our deployment process and systematise our operations. For the few remaining weeks of 2016 we began many (many!) tests in this new region, ensuring that it would be suitable for our needs. As the UK wound down for the Christmas break, this seemed like an ideal time to perform the migration. A time when our annual traffic is at its lowest and disruption would be minimised.
How did it go?
Between the 27th and 31st of December, we moved our application and web servers across to the new region, followed by our load balancers. At this point, we had our application servers hosted in London, with our database (and other data servers) hosted in Ireland. This worked well, however load times were slightly increased by this, so if you logged into your Signable account during this time, we apologise for the slower loading times!
Our database servers required more work to migrate, and unfortunately, this meant that we would have to make Signable unavailable for a period of time whilst this took place. So we scheduled it for the evening of the 30th December. This completed with a few complications, meaning the migration went on a bit longer than we had wanted it to, but this stage was complete.
It quickly became apparent that we had to perform some emergency maintenance in the early hours of the 2nd January 2017, to fully complete the migration. Which also meant that Signable had to be taken offline during this time. Again, we jumped into action, coordinating the team and ensuring that everything completed successfully, on time. By 5am on January 2nd 2017, we were serving everything from the new London AWS Region!
Has it helped?
Below is a graph which shows this period of time. This is how long our test suite takes to run, the higher line on the old infrastructure and the lower line in London. It has been cut down from 25 seconds to under 5 seconds. You should also see a similar speed increase when loading pages and our API.
Not only have we benefited from dramatically faster response times, but we have been able to ensure that the Signable infrastructure is more resilient and reliable than ever before. This puts us in good stead, considering 2016 has seen us more than triple the number of customers and more than 5x the number of monthly documents the platform processes.
So Signable is now fully hosted from within the UK. This ensures a higher level of compliance and ensures that your data remains safe and secure.