Actually here it is https://aws.amazon.com/blogs/awsforsap/automating-sap-installation-with-open-source-tools/

Yeah, there they are 🎉

This is the github link directly: https://github.com/aws-samples/aws-install-sap-with-jenkins-ansible

If you can show them some data pinpointing the problems, you are more likely to do the convincing.

I agree with @vladyslav.ukis. Absolutely.

Some of it depends also on how work and adoption happens in your org. Sustaining changes happen when teams see where their bottlenecks are. So that's necessary to be able to establish. But, sometimes it might be just for you to show where the bottlenecks are. Experience does come in handy if you want to move fast and you can work with teams to see where they need to improve and get better. Not scalable but probably fast if you are looking for immediate returns.

@dev to get to a more organic way to get better at practices, figure out a way to invest in measuring how work happens (planned, unplanned, SRE, platform, devops, interrupts, etc.). And build models to surface what's okay and what's not okay for most team. and then review with them periodically. ask them questions - "how do we get better?"

The model looks really interesting, do you have a public link to this? (we’re currently trying to codify our Engineering Practice Model with a focus on ‘how we can help the teams’ rather than ‘you shall do this’)

Currently we run a DevOps Check workshop periodically with the teams that start this conversation (“how do we get better?“). This is very much based on value stream mapping and the Accelerate Metrics (qualitatively) but it does not help the teams enough to understand what practices they can improve or adopt to mature

@dev i have received request for that. i dont have it available right away honestly but because many people have asked for this. let me get this done during the conference and share it with you all 🙂

absolutely. i will have to redact a few things. Let me get back to you all with this today.

Interested in your models @kapoor.vaidik and @florian.gysin!

Hey @vladyslav.ukis, @kapoor.vaidik, I’m happy to share what we have also. Don’t mind the format, it’s an ugly export from our Confluence where this currently lives on our end. 🙂

Currently we do the assessments for projects/teams, which is also why this assessment is not yet very streamlined and self-explanatory… I think that is a take-away for me, to go more towards self-assessment.

Cool, thanks a ton!

Here you go folks. Sharing the maturity model I covered in my talk - https://docs.google.com/spreadsheets/d/1g2iTst1wRU3WAivofnhBIfMOslqxdSNsUW6rEyJCR0w/edit#gid=1797617707.

we built it with them. so there was a lot of converge and diverge activities we did over the course of two months to lay the foundation. tech leadership, platform engg, and senior engineers from every team were asked to contribute and review pretty much everything.

along with the commitment that we will keep revisiting these things. nothing is set in stone but lets get 80% right.

Thanks - that co-creation aspect seems key to this.

I might have missed it - but how was team self-assessment completed.

were these based on what the next maturuity model step was, or based on what they wanted to prioritise in the next interation (i.e. they could choose anything from the model to meet expectations).

@slack1599 we started off with very basic assessments initially. sort of let tech leads just look at the maturity model and do a dip-stick check of where they are and what seems like the obvious next step to improvement. subsequent assessments were more structured. so teams will look at the microservices they own and where do they stand with respect to the practices being followed in those microservices

our vision was to get to a place where part of the assessment could be done in an automated fashion by inferring from the code, repo, pipeline, infrastructure (kubernetes specs) but havent got their yet

> were these based on what the next maturuity model step was, or based on what they wanted to prioritise in the next interation (i.e. they could choose anything from the model to meet expectations). slightly more nuanced. i could not cover that. and the model right now misses some of those things but we identified it. for example, you will prioritize monitoring and resilience over continuous delivery (perhaps) if the code base of a service is rarely changed but it is still in the critical path (for example auth service)

Great stuff!

Thanks!

https://docs.google.com/spreadsheets/d/1g2iTst1wRU3WAivofnhBIfMOslqxdSNsUW6rEyJCR0w/edit#gid=1797617707

Here you go folks. Our maturity model.

"prescriptive" means not just suggesting a practice at the high-level but being more specific of what we should do as an organization and how can that be done at our company... so that the teams don't get overwhelmed with (a) researching the discipline too much (2) developing the tech solutions to adopt the practices. ...hence making it easier for teams to progress

having said that, teams who understood some of these practices or wanted to deviate from the solutions (because of their context) could do so as well

I see. I guess for that you need proper buy-in to actually mandate those decisions over the entire organsiation, which is not the case in my specific case (more grassroots/bottom up)

But I see the value of at least proposing ways forward and thus limiting the decision space

yaeh so we built the model with the grass roots. representations from all the teams - leadership, platform engineering, security, sennior engineers representing all the product engineering groups

and also some junior engineers to understand the process and for knowledge dissemination

a lot of converge and diverge over the period of a month

How did you do the assessments themselves? Where they done as self-assessments by the teams? Or did you do the assessment with team-external people?

devops and platform engineering laid down the foundation but most hardening was done by reps from all the teams in many iterations until we felt this is a good place to start (or rather much better than how we used to work)

> How did you do the assessments themselves? Where they done as self-assessments by the teams? Or did you do the assessment with team-external people? self assessments initially a couple of times. for services/components that were identified critical (mostly senior engineers from multiple teams surfaced this information), they also realized that those services need to be assessed/reviewed by their tribe's architecture steering committee and their leadership.

so. the process was not fully hardened. we were iterating as we went.

Cool. I think we should also pivot more towards self-assessment, which we are currently not doing… doesn’t scale well as it is for us now.

I think that teams need to have their own work values at the lowest possible level. But that larger clusters of teams need to share some commonly agreed to principles

this makes sense! thanks

What would be a reasonable split here? Like 70% local values and 30% global ones?

I would say that the local values should easily fold into the global ones - but can be more specific and meaningful for the individual teams

This Brene Brown? https://brenebrown.com/about/

Got it thanks! https://devopsenterprise.slack.com/archives/C0150HQB6UX/p1652192562546019?thread_ts=1652192468.973739&cid=C0150HQB6UX

Haha! They were also part of the journey, and their story is worth telling. But I had to cut it somewhere 🙂

Dare to Lead by Brené Brown

1785042149

Agile Conversations, B086CYD1KW

Thanks a lot

@vmshook I spend a lot of time at surf and yoga retreats. Last summer I spent a lot of time on the beach teaching the yoga teacher about DevOps because it turns out there is a lot of similarity !! E.g. Focus & Flow, Efficiency, working together etc 🙂

ooooh! we're going to need a yoga component when we're back in session @annp - hint 😉

I just came back from a company offsite where I did a 20 minute desk yoga and guided meditation 🙂 Youtube video coming soon!!

On it!

yes!!! recognition is so important 😄 Its also worth speaking to your team, some people prefer public praise while other prefer a private recognition 🙂

yes. just have to promise to do it in a reasonable amount of time 😉

Rhetorical:grinning: question. Is there still anybody who hasn't read it?

I'll be honest.... I only read it recently 🙈 I'm so glad I did!! Also the font is quite big so it was a quick (and awesome) read 😄

Phoenix Project & Unicorn Project are my go-tos! Currently reading 'DevSecOps' by @glenn.wilson 🙂

Make Work Visible is one I go back to a bunch, too

Ooooh I haven't read it, it keeps coming up so I'll add it to the list. Thanks @vmshook

@glenn.wilson’s book is also currently on my desk!

It really vafies !!! The best I've seen is 1:65, the worst... 1:500 or more 👻

Thank you @lucasc5 great to hear you enjoyed the session 🙂