Can Ops Teams Become Agile? A Practical Kanban Journey

Introduction

In many work scenarios, operations often encounter problems that are actually related to development, quality, and testing. Operations, as the final stage of product delivery, face issues similar to those in development. The author asked whether agile Kanban could be introduced into the operations team to solve these problems.

The talk will cover five parts: challenges faced by operations, agile development methods, the operations Kanban board, the agile software lifecycle, and the conclusion that operations can also be agile.

Challenges of Operations

What can operations do within DevOps? How can the operations department collaborate with quality and development? The development path of operations includes standardization, process, and automation, but many teams wonder how to transition gracefully before full automation.

Key problems listed:

Delayed task feedback leading to perceived low efficiency.

Too many urgent tasks crowding important tasks.

Perceived low efficiency by parallel departments.

Tasks are heavily compressed and hard to quantify.

Risk introduced by a "luck" mentality.

Individual capability differences causing hidden risks.

These problems are common across departments.

Agile Overview

Operations and product delivery share similar stages; agile plays a crucial role in product delivery, prompting the question whether agile can solve operations problems.

Agile originally means flexibility; today it is recognized as flexible and elegant.

Agile Streams

Major agile streams include Lean, Scrum, Kanban, and XP.

Scrum is the most popular framework, providing a set of practices that teams can adopt quickly.

Kanban, originating from the Toyota Production System, is closely related to Lean and has grown rapidly in software.

XP, introduced in the late 1990s, contributed many engineering practices such as continuous integration, pair programming, and refactoring.

Agile Manifesto

The Agile Manifesto was created in 2001 by 17 practitioners at Snowbird, emphasizing individuals and interactions over processes and tools, working software over comprehensive documentation, customer collaboration over contract negotiation, and responding to change over following a plan.

Lean Principles

Lean originates from Toyota and focuses on five principles: value, value stream, flow, pull, and perfection.

Agile Method Selection

Method Comparison

Scrum has nine practices, XP thirteen, Kanban three; fewer practices make Kanban easier to adopt.

Scenario Analysis

In a freight telematics company with a large fleet, pure Scrum was difficult to implement; Kanban offered a gradual transformation.

Choosing Kanban

Kanban provides incremental change, suitable for multiple teams, and avoids overwhelming the organization with many new roles.

Kanban Core Practices

Visualize workflow without preset processes.

Limit Work In Progress (WIP) to control flow.

Improve flow speed gradually.

Kanban Practice Cases

Physical board for a driver platform, showing stages from backlog to completion.

Electronic board for a startup team using Scrum.

Our Operations Kanban Practice

Problem 1: Untimely task feedback

Story about a leader asking about server optimization and the lack of response.

Solution: Record tasks on a board, filter, and place them in the backlog for visibility and daily stand‑up tracking.

Problem 2: Too many urgent tasks

Example of a server removed from a cluster causing resource pressure.

Solution: Categorize tasks (virtualization, CI/CD, monitoring, automation, security, tooling) and adjust priorities daily.

Problem 3: Perceived low efficiency by other departments

Tasks can become blocked; visualizing them on a board reveals bottlenecks.

Solution: Identify blockers in daily meetings, adjust priorities, and limit each person to three concurrent tasks.

Problem 4: Tasks cannot be quantified

Long tasks hinder flow; break them into smaller, two‑week chunks.

Problem 5: Individual capability differences cause risk

Align new tasks with long‑term plans, track capacity, and ensure no one handles more than three tasks.

Software Development Lifecycle: Increasing Collaboration

Operations become a stakeholder early, contributing requirements such as logging, monitoring, SLAs, capacity planning, and security.

During construction, operations collaborate on environment management, deployment, capacity planning, and testing.

In the transition phase, operations (or SRE) handle monitoring, rollback decisions, and capacity adjustments.

Agile Delivery Three Phases

Initiation: modeling, prioritization, user story mapping.

Construction: Scrum, XP, Lean, Kanban implementation.

Deployment & Delivery: release, monitoring, and post‑deployment activities.

Operations in Each Phase

In initiation, operations provide requirements for logging, monitoring, SLAs, and capacity.

In construction, operations manage environments, deploy applications, and participate in capacity testing.

In transition, operations (or SRE) monitor production, decide on rollbacks, and manage capacity.

Summary

Operations within DevOps cover the entire post‑deployment lifecycle, acting as SRE and handling deployment issues. Applying agile thinking enables operations teams to become more flexible, efficient, and elegant.