Cloud launch process
Starting with Sourcegraph 4.0, we’re asking authors of non-trivial features launched on Sourcegraph Cloud to fill-out a questionnaire providing Cloud team with information about:
- infrastructure / resource allocation changes that need to be made by the Cloud team
- privacy / security / legal / cost implications of enabling the feature for a specific customer
- planned launch timeline
- troubleshooting and escalation steps
Non-trivial feature
For the purpose of this process, a non-trivial feature is a code / infrastructure change that fulfills one of the criteria below:
- Introduces a new component to be deployed as a container / Cloud service
- Requires changes of credentials / permissions / resources provisioned for existing workloads
- Changes Sourcegraph UI / API in a way that requires a configuration (experiment flag / environment variable / site config) change
Rationale
Two main goals of introducing this process are:
- make sure feature authors & Cloud team understand the scope of features being launched and can work together to make the feature available on time
- provide Cloud team (as default on-call responders) with sufficient context about new functionality available on cloud
Non-goals:
- adding process where it provides no value - most “features” do not require a questionnaire and can follow standard Sourcegraph launch process
- Cloud team becoming a gatekeeper of launches
Further discussion in RFC 739.
Questionnaire
Please use this as a template.
If you think it should be improved make suggestions in the doc, if you have questions, please let us know in #cloud
Past launches
- Launched Server-side Batch Changes
- To be launched RockSkip