Incident Management as a managed service
Opportunity Canvas
Problem to solve
Self-managed customers may not want to host their own incident management services because it increases the risk of downtime for the service they are using to maintain their services they provide their customers. A cloud-based IM service can be much more reliable.
One of the values of PagerDuty is that they are a HA service with many 9's of guarantee. To be suitable for enterprise on-prem we will eventually need to provide similar level of service and guarantee.
We potentially want to create a cloud-based Incident Management service that hooks into someone's self-managed installation so that they do not have to host it themselves.
Intended users
- Enterprise customers
- Self-managed customers
User experience goal
Self-managed customers can rely on GitLab for highly available and reliable incident management services.
Proposal
- Establish a cloud-based service for incident management for self-managed customers
- Establish a HA cloud-based service separate to GitLab.com for any enterprise customer - we might want to separate it from GitLab.com to ensure higher guarantee of availability (not validated)
Further details
This work supports the Incident management direction.
Questions to answer
-
- What data would we want to pass back and forth to the self-managed GitLab instance?