Since early 2017 Lightwire has been hosting 3CX instances on our geographically distributed servers, our private cloud if you will. We have been running with an N+1 design that ensured we had reserve server capacity in the event of a critical failure.
However, last month Lightwire made the call to move all new 3CX builds to AWS, and over the coming months we will be migrating all existing services to AWS as well.
This represents a fairly significant overhead for us, both in terms of hours and dollars, so why the change?
There are a number of key reasons:
- Improved failover times. As scale has increased, so has complexity. We simply can’t manage failovers well as AWS can. With this move, the time to failover in the event of a primary server outage will drop from 15-30 minutes to a couple of minutes.
- Reduced CapEx in favour of OpEx. When running our own platform, capacity planning was becoming a real headache, we had to ensure that we had spare capacity in the event of an outage while also trying to budget for new server purchases in line with forecast growth. While the monthly cost of each 3CX service deployed is now higher, it is a known overhead and one that comes with far less risk.
- Supported environment. When working with vendors, everything is easier when working within a platform they have templated models for and are able to fully understand. AWS is a fully supported 3CX environment meaning we gain ease of deployment and automation, upgrades are simplified and 3CX support is more accessible.
- Offnet platform. In the event of a Lightwire outage on a node hosting a clients 3CX server, this would result in that 3CX instance becoming unusable. Being hosted on AWS, even if a clients Lightwire internet connection is down, they could continue to use their 3CX mobile app on a 3rd party mobile network.
- Training and staff. AWS is a well-understood platform with great documentation and training material readily available. When looking to grow our team, this move will reduce training overhead. Basically, bringing people into bespoke private environments makes life hard.
We now have a good number of 3CX clients running on the AWS platform and all feedback has been really positive.
The migration of existing services will begin soon and we will handle each migration in close coordination with our client’s tech contacts and process the changes out of hours. The good news is that there is basically nothing our customers will need to do as part of these migrations other than schedule 30-60 minutes downtime while we migrate and test.
If you have any questions regarding this change, please just let us know by getting in touch via firstname.lastname@example.org.