Did you have to introduce new things/tools/stacks to accomplish that goal?
Not more than usual. We're often experimenting with and introducing new languages + tools, but there was nothing specific to this transformation in terms of languages or stacks.
In these years of transitions did you have to leave behind tools that you deemed weren't no longer serving their purpose even if perfectly valid for single tenancy? Did you get resistance on that from teams?
We did leave behind tooling + code that was only relevant for single tenancy - for example, a framework that ran rolling deploys across all tenants (ie containers in a single-tenant world) over 24 hours. There wasn't any resistance though - folks knew the problems with single tenancy, they knew it wasn't scaling with us, and there was really no one who was against the mission. The only voices of dissent were those saying we should be doing it even faster ;)
@rhymes thanks for the question.
Not more than usual. We're often experimenting with and introducing new languages + tools, but there was nothing specific to this transformation in terms of languages or stacks.
We did leave behind tooling + code that was only relevant for single tenancy - for example, a framework that ran rolling deploys across all tenants (ie containers in a single-tenant world) over 24 hours. There wasn't any resistance though - folks knew the problems with single tenancy, they knew it wasn't scaling with us, and there was really no one who was against the mission. The only voices of dissent were those saying we should be doing it even faster ;)
Ahah nice to know, thanks for the reply!