Hi everyone!
I really appreciate the feedback here.
It's true, the project is 5+ years old.
The industry was very different back then, we were barely starting to think about containers at scale. The cloud was a thing for sure, but the main contributors were mostly working in places with physical infrastructure.
Over the years we've discussed how we can reboot the project. There is always interest and desire, but time is hard to find :)
The project will stay up as it is because we still feel there is value in learning the basics.
Over time the understanding of what is "basics" is evolving.
This year I'm chairing the SREcon conference in Singapore.
The program has a "Core Principles" track which talks about things like deep dives on linux memory management, database indices, how BGP works, etc.
There is a lot of desire to still have these super deep dives into underlying technology.
I hope we can find a way to turn this into a stronger curriculum with labs and teaching exercises one day :)
I just wanted to say that I've found the OpsSchool documentation here valuable, and I still think it has a lot of value. I would encourage you to reboot and update it. I've had the site bookmarked for a long time and it's one I've returned to many times over the years. Back in the day, the way a sysadmin (now "DevOps engineer") learned all this information was by reading the paper manuals which sometimes came in binders.
I feel like there is still a lot of value in "deep dive" information no matter how pervasive automation becomes in our industry.
As someone who has been practicing a lot in an 'old school' environment (many of our customers still run our software the old way) I honestly think this still is a very good curriculum.
Actually I think many of the things listed, people underlook at them. I've seen with my eyes digging very deep on "cloud" and "devops" stuff fail completely (and spectacularly, I must admit) ad very basic things.
I'll be digging deeper into opschool.org, is there a way to have an offline copy (besides running httrack) ?
(edit: looks like, as per a comment further down, a lot of this is five-ish years old; the decisions made in its writing/presentation make a lot more sense in that light. Leaving comment for posterity.)
So I get the overall goal here, but reading this, this feels like somebody's trying to fight the last war.
> Since the early 90’s, operations engineers have been in high demand. As a result, these positions often offer high salaries and long term job security.
...because most of the control surfaces offered minimal automation and there were many fewer options for management. This isn't even a "well, AWS/GCP/Azure will eat your lunch" thing, it's a "the robots can do most of this already" thing. There are still roles for network engineers, large-scale datacenter management, and that sort of thing--but this "curriculum" reads as very...hands-on...when a holistic, automation-first approach is overwhelmingly preferred in industry (outside of the trailing edge, where those high salaries and that stability aren't nearly as common).
Again: not saying "use the cloud, datacenters don't matter." But I'm not sure that this encompasses what "ops" typically means (as opposed to "IT", which is itself becoming a largely automatable and automated function, or "devops", which will get there sooner than later) and I'm very unsure that places that will still be dealing with data centers in ten years, instead of outsourcing it, will benefit much from juniors with this skillset instead of one built around software definition and a pervasively automation-first mindset.
Not totally off-base, but any time I'm feeling anxious about the value of "old" skills, I realize that might just be SV talking. These "old school" skills will be useful for some time to come in less-cutting-edge orgs, labs, etc.
courses like this tend to be less of a function of what would be the most useful for the profession as a whole and more of a function of what someone would need to know to get a job. these skills are definitely things people would need to know to get an SRE/DevOps/Ops/IT job in 2019. (I understand that those 4 jobs are not the same.)
Wow, this is really great! Soft Skills 201 could be a book unto itself, nice to see this stuff being discussed.
In general, I think the soft skills sections should get a lot more promotion (it's buried so far down the sidebar I had to scroll!). The technical stuff is already well-documented elsewhere, but nobody ever told me how important being politically savvy was in this niche until I had already blown both feet off several times.
This lacks the most critical part of ops in designing systems that are repeatable and debugable, discrete build steps on a time line. You can’t reason about any complex system without some core logic and this is most absent in operations.
[+] [-] avleenvig|7 years ago|reply
It's true, the project is 5+ years old. The industry was very different back then, we were barely starting to think about containers at scale. The cloud was a thing for sure, but the main contributors were mostly working in places with physical infrastructure.
Over the years we've discussed how we can reboot the project. There is always interest and desire, but time is hard to find :) The project will stay up as it is because we still feel there is value in learning the basics.
Over time the understanding of what is "basics" is evolving. This year I'm chairing the SREcon conference in Singapore. The program has a "Core Principles" track which talks about things like deep dives on linux memory management, database indices, how BGP works, etc. There is a lot of desire to still have these super deep dives into underlying technology. I hope we can find a way to turn this into a stronger curriculum with labs and teaching exercises one day :)
- One of the original OpsSchool founders
[+] [-] wyclif|7 years ago|reply
I feel like there is still a lot of value in "deep dive" information no matter how pervasive automation becomes in our industry.
[+] [-] znpy|7 years ago|reply
As someone who has been practicing a lot in an 'old school' environment (many of our customers still run our software the old way) I honestly think this still is a very good curriculum.
Actually I think many of the things listed, people underlook at them. I've seen with my eyes digging very deep on "cloud" and "devops" stuff fail completely (and spectacularly, I must admit) ad very basic things.
I'll be digging deeper into opschool.org, is there a way to have an offline copy (besides running httrack) ?
[+] [-] kim0|7 years ago|reply
[+] [-] eropple|7 years ago|reply
So I get the overall goal here, but reading this, this feels like somebody's trying to fight the last war.
> Since the early 90’s, operations engineers have been in high demand. As a result, these positions often offer high salaries and long term job security.
...because most of the control surfaces offered minimal automation and there were many fewer options for management. This isn't even a "well, AWS/GCP/Azure will eat your lunch" thing, it's a "the robots can do most of this already" thing. There are still roles for network engineers, large-scale datacenter management, and that sort of thing--but this "curriculum" reads as very...hands-on...when a holistic, automation-first approach is overwhelmingly preferred in industry (outside of the trailing edge, where those high salaries and that stability aren't nearly as common).
Again: not saying "use the cloud, datacenters don't matter." But I'm not sure that this encompasses what "ops" typically means (as opposed to "IT", which is itself becoming a largely automatable and automated function, or "devops", which will get there sooner than later) and I'm very unsure that places that will still be dealing with data centers in ten years, instead of outsourcing it, will benefit much from juniors with this skillset instead of one built around software definition and a pervasively automation-first mindset.
[+] [-] rconti|7 years ago|reply
[+] [-] femiagbabiaka|7 years ago|reply
[+] [-] gowld|7 years ago|reply
[+] [-] stoic|7 years ago|reply
In general, I think the soft skills sections should get a lot more promotion (it's buried so far down the sidebar I had to scroll!). The technical stuff is already well-documented elsewhere, but nobody ever told me how important being politically savvy was in this niche until I had already blown both feet off several times.
[+] [-] argd678|7 years ago|reply
[+] [-] leetrout|7 years ago|reply
[+] [-] incomplete|7 years ago|reply
[+] [-] avleenvig|7 years ago|reply
[+] [-] saifulwebid|7 years ago|reply
Because no operations engineer should use MS Windows? :see_no_evil:
[+] [-] mverwijs|7 years ago|reply
[+] [-] nerdkid93|7 years ago|reply