The benefits of postmortems

The benefits of postmortems

Postmortems are a well established process followed in the aftermath of an incident. Often the most visible output from the process is a structured document, with some or all of the following components:

  • Overview of what happened
  • The impact
  • What triggered the incident
  • When/how did we detect the issue
  • How we fixed it
  • Timeline
  • Root cause analysis
  • Learnings
  • Actions
  • Internal email
  • External message

That’s all well and good, but in my opinion not the primary reason we do postmortems.

The real benefits of a postmortem process are the discussions we have following the incident, and the follow up investigations and investment work. The process forces a group of people to look critically at the incident, identify what should have been better, and make decisions on how we can improve the system to prevent future incidents. We then follow that up with clear and well-defined actions that are prioritised at the very top of our backlog.

A postmortem process is great to have. It gives you the time and space to have these discussions. And the document does help to capture and share knowledge and is certainly worth doing. But it’s important to remember what the primary objective is, and it isn’t to write a document.

---

Originally published at https://andrew-jones.com/blog/the-benefits-of-postmortems/

Cover image from Unsplash.

要查看或添加评论,请登录

Andrew Jones的更多文章

  • A Data Platform is not just for analytics

    A Data Platform is not just for analytics

    Data is at the heart of every meaningful service, and it’s the effective use of that data which builds a product, and a…

  • Data Contracts

    Data Contracts

    Almost all data platforms start with a change data capture (CDC) service to extract data from an organisations…

  • We had an incident, and it was great

    We had an incident, and it was great

    We recently had an incident with our data pipeline, resulting in data being lost on route to our data platform. Of…

  • The democratisation of Data Science

    The democratisation of Data Science

    There’s a trend in the industry to make data science and machine learning more accessible, allowing engineers to build…

  • What does a Tech Lead do?

    What does a Tech Lead do?

    I've been a Tech Lead for a few years now, though I'd say I've only been a good Tech Lead for about a year. So what…

  • Lambda Architecture in 2020

    Lambda Architecture in 2020

    As I start to think about some of the upcoming projects we'll be working on over the next year and how we might go…

  • You build it, you run it - you manage your data

    You build it, you run it - you manage your data

    Many tech companies follow the "you build it, you run it" mantra - and for good reason. You want to empower your teams…

  • If your data is worth storing, it's worth structuring

    If your data is worth storing, it's worth structuring

    When some people talk about a Data Lake (or Hadoop, or even just Big Data), they go on to say that we can store all our…

  • Momentum

    Momentum

    In the Michael Connelly books, Detective Harry Bosch often talks about the importance of momentum when working a case…

  • Why your master datastore should be immutable

    Why your master datastore should be immutable

    Thinking about how your master dataset is stored and managed is arguably the most important part of architecting a…

社区洞察

其他会员也浏览了