We’ve been busy over the Christmas period and are excited to announce that Action Links is now available in North America!
Significant infrastructure developments
The Action Links architecture has always included a range of exciting technologies, from cutting-edge tools like serverless functions to NoSQL databases. However, to optimise the end user experience and reduce the latency between our hosting and the products we integrate with, much of this infrastructure has to-date been operated from data centres in Europe.
To enable us to expand our offering, we've undertaken a significant overhaul of the Action Links architecture, bringing a range of benefits and allowing us to open up the product to our friends overseas.
Benefits of the new architecture
We’re really excited to be able to offer our services to prospective clients in the US and Canada, but there are also other beneficial outcomes of the work:
- Load balancing: We’ve introduced additional load balancers to the architecture. In addition to enabling us to separate regional traffic, they also give us more granular control over how requests are routed to back-end services – important for future scalability and performance monitoring.
- Serverless: We’ve migrated from serverless functions to serveless containers, allowing for more efficient code (that should make future development easier), and resulting in fewer cold starts (where a service takes a little while to boot up).
- Caching: We’ve added a new caching layer that sits in front of the services that handle Action requests, served by a Content Delivery Network that caches responses ‘at the edge’ (i.e. in multiple locations around the world). This layer allows us to configure caching behaviour for each Action individually, and in the future will enable us to give clients more control over how their Action Links handle caching.
- Microservices: We’ve got more control over our microservice architecture, adding extra redundancy by enabling us to share the load more easily between services as we scale. It’s also made it possible for us to deliver our new Pre-Action layer, and in the future, our Post-Action layer too.
Retained benefits from the existing architecture
In addition to the benefits of the new architecture, we've also retained the firm ground we had before, including:
- Greener data centres: We use Google Cloud’s very efficient infrastructure and have chosen to use specific data centres that have a lower carbon footprint. The energy sustainability of these data centres is measured in CFE (% of carbon free energy used) and the carbon intensity of the grid in that country (i.e. how much carbon is produced for each kWh of electricity supplied). In particular, we currently use eu-west2 (London) with a CFE of 85% and grid carbon intensity of 166gCO2eq/kWh, and us-central1 (Iowa) with a CFE of 92% and grid carbon intensity of 445gCO2eq/kWh (sadly the USA still has some way to go to reduce the carbon intensity of their electricity grid!).
- Greener architecture: We have a serverless back-end that consists of multiple serverless microservices. Whereas traditional servers are on all the time, our services run on shared servers that only ‘boot’ and scale up in response to requests, which means our servers aren't wasting electricity. Electricity is also consumed when transporting data between data centres, and from a data centre to the end user. By considering the location of our data centres in relation to the location of the services we integrate with and that of our end users, we’re able to reduce the distance the data has to travel (and provide faster load times for users!).
- Security: In addition to multi-level encryption for your integration keys, Action Links benefits from Google Cloud’s added redundancy, and protection from DDoS (Deliberate Denial of Service) attacks.
To find out how you can use Action Links to empower your customers, book a demo or get in touch.
