Incident Response Lifecycle

The Incident Response Lifecyle working group is intended to document a shared incident response protocol and knowledgebase.

Attributes

Property Value
Date Created February 01, 2022
Target End Date April 30, 2022
Slack wg-incident-respose-management-framework
Google Doc Incident Response Management Working Group (internal)
Issue Label WG-IRM (gitlab-com/-org)

Business Goal

  1. Increase efficiency through common incident response, analysis, documentation, ongoing management and reporting methods.
  2. Increase transparency through improving visibility and communication of incidents to business and e-group
  3. Support results by building our clients’ confidence in GitLab’s ability to quickly resolve and communicate incidents when they occur
  4. Align Incident Management activities and priorities with those of the business
  5. Prepare materials for the creation of training modules for the Engineer Department on the Incident Management Process at GitLab
  6. Highlight dogfooding opportunities

Exit Criteria

  1. Single source of truth documenting incident response management that will be applicable to all areas of Engineering and teams who provide Incident Response
    • Each functional area of Engineering will develop their own Incident management requirements for identifying and reacting to service outages or security threats.
  2. Create a comprehensive knowledge base for GitLab team members to help them understand how incident response teams implement the IR process

Outcome

  1. Help teams across GitLab lower MTTR

Other Investigations

  1. Improvements in feedback and learnings from Incident to build resiliance
  2. Service Catalog

What do other companies do?

How is IR done today?

  1. SIRT
  2. On-call
  3. Reliability
  4. Support

Noted issues

Roles and Responsibilities

Working Group Role Person Title
Facilitator Anna Liisa Moter Manager Reliability
Exec Sponsor Steve Loyd VP Infrastructure
Member Anthony Fappiano Manager Reliability
Development Functional Lead Dan Croft Senior Engineering Manager, Ops
Member Sam Goldstein Director of Engineering, Ops
Member (CMOC) Kenneth Chu Support team
Member Kevin Chu Group Manager of Product, Monitor

Requirements and Considerations

Actors

  • Reliability Engineers
  • SIRT Engineers
  • Development Team
  • Quality Team
  • Support Team

General

  • As a GitLab team member who can raise an incident, I know how incident can be initiated
  • As a GitLab team member who can raise an incident, I have a general understanding about incident severity levels
  • As a GitLab team member who can raise an incident, I understand the high level process of Incident Management and its importance to the business
  • As a GitLab team member who can raise an incident, I can contact the right team via dedicated slack channel.
  • As a GitLab team member who can raise an incident, I can easily find a page in the handbook that documents the Incident Response Procedures

SIRT Engineer

  • As a SIRT Engineer I know how to pull relevant resources from other teams when I need assistance
  • As a SIRT Engineer I can easily categorized the incident
  • As a SIRT Engineer I can identify triggers and indicators
  • As a SIRT Engineer I know where to document the incident details
  • As a SIRT Engineer I know when to transitions from Incident identification, to mitigation, to remediation, and post to incident activities
  • As a SIRT Engineer I can follow a reporting process to handoff incidents, or provide updates to Management

Reliability Engineers

  • As a Reliability Engineer, I know how to level an incident in a manner that is consistent across the org
  • As a Reliability Engineer, I know how to engage the other roles during an incident
  • As a Reliability Engineer, I know when to transition from Incident identification, to mitigation to resolution and post-incident activities

Development Team

  • As a leader in Development who is part of the Incident Manager rotation, I am clear on the role’s responsibilities and how the role supports the Incident Management process.

Quality Teams

Support Team

  • As a Support Engineer, I know how to create a status page
  • As a Support Engineer, I know the differences between the Incident Status states on the status page
  • As a Support Engineer, I know how frequently to update the status page
  • As a Support Engineer, I know how to engage the Incident Manager or EOC when asking about feedback for an update I am about to post on the status page
  • As a Support Engineer, I know how notify the stakeholders
  • As a Support Engineer, I know how to find related tickets in Zendesk and the GitLab issue tracker to help access the impact of an incident
  • As a Support Engineer, I know how to contact users if their usage of GitLab SaaS was restricted due to an incident
Last modified March 26, 2024: Update links to remove redirects (5b495046)