UpskillingSRE

Because it puts the Ops in DevOps!

Go Cloud/Cloud native is new normal now.its where traditional ops can't match speed with such modern tech. also SRE is highly sociotechnical discipline hence SRE gaining traction across the industry.

Organizations looking to achieve higher levels of velocity understand the importance of wiping out toil, automating manual work, and breaking down the boundaries between silos

Because we have always had a Dev "team" member who worked with Ops teams who did this job. Now it is becoming formalized. It has always been important.

It brings How into Picture of DevOps Practice implementation? #DevOps #SRE

Our systems are growing more complex.

In the words of @tammybutow, “Reliability is feature zero”. SRE allows us to create velocity and resilience that is safer, known and consistently measured.

Reliability and Resilience are not new concepts or needs. What is driving 'SRE' is the need to rethink Reliability and Resilience in a Cloud world, where you are not able to optimize the underlying infrastructure to custom-fit your application's needs.

Because it's a new title? That is only a semi-gest! Some people I talk to think it's purely another form of maintainability/availability/security type work I think it's important to distinguish anything particularly unique to these facets of developing & managing tech

I'd say it is more about the response of operations to the rapid and dramatic changes that have occurred in software delivery in the past decade.

it applies the software engineering mindset to system administration domain. something which was missing big time.

Culture, DevOps and SRE go hand-in-hand, it should be ongoing, rather than emerging.

It helps in building reliable resilient complex software systems with predictability

imo we've had SRE forever, we just used to call it 'ops'. devops is great, but i think the lived experience of teams is that you need people who can focus on foundational work.

Provides a different view on speed, security, availability and integrates devops

I think SRE is emerging because of the increased need for stable, secure systems and changes in development processes (moving faster)

The size, scale and speed of modern software coupled with our dependence on technology as a society has fostered the need for greater focus on increased quality of service delivery.

@felidue When we say different view - what is the perspective. How is it looked at differently? These are the questions I get asked and not always sure how to answer.

my initial feel is to say that SRE is about finding the balance in a 'rocking boat'. If we're to navigate someone standing and being an idiot in it or the waves sideways we need to handle that & more.

We need a new approach to Operations that aligns with Agile and CI/CD and DevOps

@simonejomoore It's viewing infrastructure issues as software problems. Treating the infrastructure like software as well, versioning configs, etc.

@JeremyCMorgan I think we need to be clearer on the basic 'how' of this with customers (the non-tech types). It's still anathema to some since infrastructure is physical still for many

@JayneGroll Still so much gap sometimes of where is the Ops in DevOps. I can't understand since I used to include even my service desk teams in design/build/release teams

@BealHelen ops is part of devops, isn't it?

Wondering if you have time for a 2nd question as this one is on fire.

@felidue Totally! But I've had more than one client say to me that DevOps is about dev and there was that whole NoOps hiatus years ago... SRE helps bolster the ops position IMHO

@simonejomoore great point, much of SRE isn't possible on prem, but there are pieces you can take from it.

@simonejomoore SRE endeavors to bridge the gap between the operations team and development team. Where the development team may focus on fast production, the operations team may focus on measures to prevent production blowing-up. SRE comes in to harmonize the two different views.

DevOps and SRE can overlap with SRE, as a role, focusing on improving reliability, while DevOps focuses on building the culture and mindset required to speed up the entire SDLC.

DevOps is about maximizing the efficiency of your application delivery pipeline. SRE is about maximizing the Reliability and Resilience of the applications and services deployed in production. Both are cultural movements

(edited)

There are some SRE specific practices around toil and SLOs and error budgeting that are unique to SRE but extremely valuable to DevOps practitioners

Devops aims at unifying software development and software operations while SRE happens when a software engineer is tasked with operational tasks

How do we balance change velocity vs. stability, reliability and other operational attributes? How do we improve performance of the system? How do we avoid incidence response burnout? SRE is the answer there complimenting DevOps.

Both cultural movement and sociotechnical discipline one to make delivery faster other helps to make reliable

Solid answers friends!

There is a lot of overlap, and they are mutually beneficial. Proper SRE focuses on reliability and resiliency, allowing developers to know they can push features with confidence.

its not apple vs orange comparison. They are more of close colleagues structured in a manner in order to break down silos in order to deliver better software faster @DEVOPSINST

SRE is the last puzzle piece to the "continuous" paradigm from Agile, DevOps and SRE - continuous operations!

@mbiswajit Chaos Engineering a subdiscipline of Resilience Engineering is a good methodology for balancing stability, reliability and performance.

SRE and DevOps are essential the same conversation "under the covers". Its the same intellectual and practical roots, with SRE taking a very Ops point of view (and gives more prescriptive advice)

@anuragthoughts I think the 'socio' is important here. Technology is a catalyst but people interaction with the tech has shifted, is shifting and continuing to shift. Humans are not evolving as fast as the tech yet… ooo this is a fun tangent.

Knowing your Reliability (Availability/Performance/etc) needs

Visibility into measurement

Availability with highest Quality

communication! who are your users, primarily? i'd say they're other developers. talk to them, help figure out where the pain points are, and work together to make better tools and foundational practices.

The marriage of engineering and service management

I will say Funding funding and funding. Once you have funds you grow and You can only realize SRE as you grow else traditional IT can deliver your BAUs operations.

Measurement of Qulaity of Services (QoS) parameters

SREs ensures that everyone in the company agrees on how to measure reliability

If you start thinking about microservices vs. monolithic - the answer to this is more obvious. Having someone on the team to help track a decomposed application is critical.

Curiosity. Environments are increasingly complex and practices are constantly evolving and so a constant desire to learn, ask questions, and embrace the unfamiliar is needed.

Know you are going to get burned. Working with with improbable problems for the first time, there is no play book for the unknown. Keep your head up and teach your learnings :-)

Service Level Objectives reinforced with consequences (what do we do if the SLO is broken)... if you are doing that you are on the path to SRE. Sounds way easier than it actually is!

customer experience

SRE satisfies DevOps principles

Visibility. You don't know what you can't see. No amount of insight into operations is too much.

proactive monitoring on thresholds (scalability, reliability, performance)

Comes back to my question on how it fits in the value stream and measurements that meet that outcome.