Team Radia Sprint Ceremonies

We currently use the Infrastructure board to track our current sprint, as well as the issues in our backlog.

Our Sprint Principles

We work in sprints, because doing so:

  • Helps us focus on burning down the backlog
  • Allows us to understanding the velocity we’re moving at

We agree that:

  • We do not work on issues outside the sprint or add new issues to the current sprint, unless you are Primary or Secondary On-Call
  • We strive to only work on one issue at a time, so the In Progress column always describes what we are working on and standups are more efficient


These events are shared within the shared calendar

  1.Mon 1.Tue 1.Wed 1.Thu 1.Fri 2.Mon 2.Tue 2.Wed 2.Thu 2.Fri
Weekly Omnibus X         X        
Video Standup     X         X    
Sprint Planning X                  
Backlog Refinement           X        
Strategy       X            
Wargames                 X  


In whichever format they take place, standups should answer the following, on a per-person basis:

  • Y: What I did yesterday
  • T: What I’m doing today
  • B: Any blockers I have

As much as possible, the work that each team member reports should match up with the issues they currently have on the project board.

Additional discussion, whether in general or regarding specific issues, should only occur when each team member has addressed each of these items.

Video Standup

Standup begins once every team member has joined the call, or 5 minutes after start time, whichever occurs first. Each person on the call reports their yesterday / today / blockers described above. Those who are unable to make the call should report their standup information in Slack as close to the standup meeting time as possible.

During a video standup, please observe these courtesies:

  • Constrain standup time to ~15-20m total, or 1-3m per person.
  • At the start of a standup, there typically should be only one In Progress issue per person.
  • Avoid specifics/lower-level details where they are not needed
    • A question or two is reasonable but if we go past that, say “let’s Parking Lot this”, and continue discussion when the standup is over.
  • The standup meeting will not be a deep review like either of the Monday sprint meetings.
    • The exception is when reviewing blocked items in order to help unblock team work; however, please Parking Lot deep dives into these items until the end of the standup

Once all team members have finished reporting their standup items, a verbal confirmation that “standup is now finished” should be declared, allowing team members to drop off the call if they have other projects to work on, or stay on if they’d like to be involved in Parking Lot/miscellaneous discussions

Per-Sprint Meetings

These occur once per sprint (except for the Omnibus, which is the first meeting of each week).

Weekly Omnibus

This acts as a planning and sync point for the team to talk about priorities for the next 2 weeks.

Sprint Retro

(Part of Weekly Omnibus; done at the start of a new sprint)

Endeavor to understand what worked well in our previous sprint, what didn’t go so well, and how we can improve for the next sprint. Use the Google Doc to track the discussion, and

  1. Vote on items to be addressed and document discussion as much as possible
  2. Create issues/action items when possible

Sprint Planning

(Part of Weekly Omnibus; done at the start of a new sprint)

  1. Team members take a few minutes to update their cards on the project board, depending upon what work was completed in the previous sprint, and what work they plan to do during the current sprint.
  2. Issues should be pulled from the top of the Backlog if a team member has no issues within the To Do / In Progress columns.
  3. Focus should be on issues brought up during the On-Call review (i.e. “what is happening this week”) and the product priorities defined by the Product Owner.

Backlog Refinement

(Part of Weekly Omnibus; done midway through the current sprint)

  1. Arrange issues in the backlog by priority
  2. Add detail to upcoming stories that need it
  3. Split large issues into smaller issues

To preserve velocity, issues within the backlog should ONLY be added to the current sprint if they can be completed before the end of the sprint.


Open-ish session for frequently-mentioned topics which are better discussed in a video/standup/meeting format, rather than in Slack/text. Use the Google Doc to drive the meeting; one team member should act as a scribe and, as much as possible, record all ideas/concepts brought forth for discussion.

  1. Add topics to the top section of the doc before/at the start of the meeting.
  2. Once meeting begins, vote on topics to be discussed by adding a + mark at the beginning of the line for each topic you’re interested in discussing. Please limit yourself to 1 vote per topic.
  3. After voting is complete, discuss whichever item(s) have the highest number of votes, in order of 1) number of votes, then 2) order in the list. All of these items should be moved into the current-date section of the doc and removed from the top list.
  4. During/after the discussion, create action items (i.e. GitHub issues) required for anything we want to move forward on.
  5. Any topics that were moved from the top list, but were NOT discussed in the meeting, should be moved back to the top of the doc. All votes should be removed from the top-section topic list.

During the discussion, each team member is empowered to:

  • Encourage buy-in / consensus from other team members
  • Recognize when a discussion begins to trail off into bike-shedding, and suggest pulling back to the main topic



Toil is used to describe work that does not materially affect production, or the way the team operates, but is (often) essential work that leads to additional tech debt if it is not addressed. It is thusly named as the work:value ratio is further skewed towards the work side than the value side.

  1. The Primary On-Call is the only team member that handles Toil during a given week, and should hand off any relevant issues to the next On-Call during the Weekly Omnibus.
  2. When necessary, Toil should be discussed in Video Standups after the standup reporting is completed.
  3. If an issue has been gaining scope creep, it should be scoped into the following sprint and/or broken into smaller issues.


“A strange game. The only winning move is not to play.” - WOPR engineers engage in bi-weekly contingency planning exercises in the form of role playing war games to improve operational readiness and share knowledge.

This is part of our overall set of contingency planning (CP) practices. (See IT Security Procedural Guide: Contingency Planning (CP) CIO-IT Security-06-29 for CP requirements and NIST Contingency Planning Control Family for the set or related controls.)


  • Dungeon Master (DM) - On-Call Engineers (primary from 1st week of sprint and primary from 2nd week)
  • Player Character (PC) - Everyone else. Collectively “the party”.

Prep: Each DM must prepare prior to the wargame:

  • Devise a small scenario - Remember that time is limited! (Previous post-mortem notes are full of potential ideas.)
  • Apply to your personal sandbox
  • Test the scenario
  • Test the fix
  • Note the break and fix (WITHOUT SECRETS) in a GitHub Gist
  • Share the Gist with your fellow DM

Pre-Game Communication:

  • Prior to starting the wargames, one of the DMs will send the following announcement in the #login-situation channel: (Replace {LIST} with the environments testing will be conducted in.)
    @login-devops is commencing war game activities in the following environments: {LIST}
    Disregard alerts from these environment(s) till further notice.


  • All members converge in the situation war room (See Meet Link)
  • The DM starts by telling a brief story guiding the party into a starting point. (Example: “You just deployed to the yourname environment and now PIVCAC logins are failing…”) At a minimum the DM must provide:
    • The environment name
    • One or more symptoms to diagnose
  • The PCs can ask questions, but are mostly self guided.
  • The game runs for 15 minutes or until the party restores service
  • The alternate DM (one not running the game) acts as referee

Post-Game Communication:

  • After completion of all wargames for the day, one of the DM will send the following announcement in the #login-situation channel:
    @login-devops has concluded war game activities in the following environments: <LIST>

Post Game: