I did some freelancing cloud work recently and was shocked at what I saw out there. From small one person operations to startups to large multinational Fortune 500 corporations I saw the same pattern repeated over and over. People using the cloud to spin up infrastructure that have no experience building out infrastructure and making really bad and dangerous mistakes.<p>Amazon, Google, Microsoft and others make it simple to get going but the devil is in the details when it comes to building it right, safe and ready for production. Just a few examples:<p>• Putting RDS SQL Servers on a public IP with no protection<p>• No templating of servers so if it disappears you have no record of how to get a new one running again<p>• Servers with SSH password authentication turned on with passwords of “Password”, because SSH key auth was “too hard for our devs”<p>• No backups because “it’s in the cloud, isn’t the cloud backing it up for me, like iCloud?”<p>I have an AWS Solution Architect Professional certification and agree you could get a cert and still not be qualified to run or design much on AWS but it does put you ahead of much of what I’ve seen out there.
The question: "Why it’s so hard to find roles in cloud technology, while jobs go unfilled."<p>The answer: "The key thing here is that businesses are not hiring fresh-out-of-certification AWS experts to take over the infrastructure of their established operations teams. They want people with solid experience, a risk-free bet that the new hire can execute the tech flawlessly. This often means insane job requirements [...] and an existing track-record of success."<p>So jobs openings stay empty because companies have unrealistic expectations. Is that something new and surprising? This is the pattern in tech recruiting as far back as I can remember.
"The jobs opening up include a requirement to know all these skills in addition to everything else."<p>I've been casually looking at job postings and sometimes wonder what drugs the job poster is on when writing job requirements.
Every year cloud providers announce tens of new services, poorly tested and far from production ready. Non-ops eng teams and management are always happy to use the latest buzzword out of AWS, then ask their ops team to make it somehow work.<p>Certifications are moot. I don't care if you are certified, or if you have used a cloud (or kubernetes) following a tutorial. I do care if you have seen enough issues throughout your career, so you can debug every time (not when, nor if) mud hits the fan. I don't care if you can type a port in a yaml file or in your browser. I do care if you know how many ports are there, what's the ephemeral port range and so on. Systems are 75% experience, 25% basic knowledge, you can't get away just with certifications which matter only after you get experience and knowledge.<p>If you are experienced in ops, I know you can learn and work out any cloud. If you claim you are experienced with a specific cloud though, then better have to share a ton of war stories or it will just be a huge red flag on your CV.<p>Let's call it for what it is, clouds, like data-science are hot fields right now. It's only natural that many people will go after a good-paying job claiming expertise in the field.
"The Creeping IT Apoclypse" is on the same theme:<p><a href="https://news.ycombinator.com/item?id=18930781" rel="nofollow">https://news.ycombinator.com/item?id=18930781</a>
"Hi! I'm your new cloud outsourcing specialist. I'm here to take your old IT structure and those servers in the closet devils-you-know with something so intangible and 'elsewhere', when it stops working some day you cannot possibly send an army to the closet to fix the problem. You can all just wallow in your helplessness and go home and wait for a text message. When do I start??"
Just partially related to the article, but mentioned in it nonetheless, I just wanted to ask if “serverless” is now the new hype-word instead of containers/Docker/K8? I’m curious because I’ve started reading about it in different contexts more and more lately but afaik it doesn’t come with any clear advantages over the “classical” way of doing stuff.
<i>Typically we have shortening 5-year cycles in tech that cause clamoring for new skills. First it was the internet (“HTML! Perl! PHP!”) then mobile apps (“Android! Objective C! React Native! Flutter!”) and now cloud (“AWS Certification! DevOps! SecOps!”). IoT and machine learning are probably next. The technology itself doesn’t really matter — companies just want an army of tactical workers to build this stuff tirelessly.</i><p>It's Kubernetes and microservices! Even though Kubernetes is what, four years old, where I am, numerous employers are chomping at the bit and demanding that you be an expert at these. Now, I'm a crusty old Linux sysadmin/programmer who puts "SRE" and "DevOps" on his resume, and I'm noticeably getting spurned because I don't have "5 years experience" with "containers".<p>At first I was worried that my skills were obsoleted, having worked in large enterprise environments doing infrastructure engineering for over 15 years. I worried that I was too "systemsy" and not enough "cloud-y". Oh, and even though I spent years using TeamCity and Jenkins, I was also spurned because I didn't use the right lingo within "CI/CD pipelines" and demonstrate a Docker deployment system <i>exactly like they used.</i><p>Then I realized something.<p>The job descriptions are asking for <i>five years experience total</i>, number one, and the second red flag: the salaries and contract rates are <i>lower</i> than the rates for large enterprise-y type jobs (minus management).<p>Conclusion: hipster tech coincides with youth and coincides with companies following trends and buzzwords over wisdom and experience.<p>Back to the point. Someone with a strong basis in the core disciplines of systems design, infrastructure automation, perimeter security and so forth are going to be able to adapt to AWS knowledge, as I have (for the record I have two AWS Associate certs and didn't find them hard). And as the author points out, there are these 5 year cycles where all we look for is experience in certain tech. Is it any wonder that there are horror stories about fragile and insecure cloud environments?<p>It goes back to the incompetence of the hiring process which we see again and again discussed on HN.
<i>> "10 years serverless" as a hard requirement...</i><p>So what we have here is a failure in the hiring process. The people writing the jobs don't actually know the space, but are looking for candidates with literally impossible amounts of experience.
That S-Curve on cloud seems off.<p>Most Federal agencies are on or moving to Cloud services right now. They are squarely toward the top right of the S.<p>I don’t think you can say only “large trailblazers” are there. We can argue about how many services have moved and how it’s being used but the whole “let someone else manage it” argument is long done and normalized.
29 year IT veteran here. I don't agree with a lot of this article.<p>I remember panic-talk when outsourcing began. When offshoring arose. Etc. None of these extinguished the need for IT workers.<p>I view the cloud as the next generation of the mainframe. By that, I mean your experience and understanding of the environment (nobody will master all of it) will be a career differentiator. Knowledge of Kubernetes, Istio, etc. will help you to navigate the next career paths. System admin skills (and associated extras, like networking) will likewise move you up the ladder.<p>Bill Gates once said there is an infinite number of problems yet to be solved, an infinite amount of work to be done. I believe him.
Disclaimer: ex preferred partner early AWS client-facing Fortune 200 consultant.<p>The senior sysadmin at the DMDC around mid California coast: only one certification, VMware. A French offensive exploits broker, solid researcher and reeigne: no certs. I only have a few certs because work paid tens of thousands for offsite training. A few certs are okay because corporate HR wants staff to demonstrate continual learning and achievement, lots of low-quality ones (A+) are a negative signal. It's more important to keep up on news, tools and techniques on the front lines.<p>My opinion is managing \<i>aaS needs solid rack&stack ops experience at scale. For IT folks: either start huge if you can or start small and move towards huge. The biggest challenges are customer communication, managing expectations, anticipating and planning where possible and showing empathy... it's a 5% technical skill and 95% getting along productively game. Communicate the business impact, don't overwhelm customers with </i>unnecessary* detail.
I really appreciate this part:<p><pre><code> Some of these firms realize they need to renovate their IT
but they attempt to do this through hiring “Digital
Transformation” roles which do little but frustrate the
revolving door of hires because they have no executive buy-
in and budget.
</code></pre>
This kind of thing makes it easier for me to hire. My org swallowed the risk and moved our production infrastructure entirely to k8s. I've no expectation that I'm going to be able to hire engineers with years of experience with that, but my team has already cultivated those skills. Aside from all the benefits and pains of being on k8s has, it makes it easy to attract great engineers with lots of related experience looking to leave a company where they've heard they need to adopt something and fail to put the resources behind it.
This is an age-old problem in IT. Some new technology and/or fad comes along, and bunches of organizations suddenly want an <i>expert</i> in it.<p>It's impossible to find enough experienced in the new field because it's so new and/or growing too fast. But the HR department or technically-inexperienced managers don't understand this and complain to politicians et. al. of a "shortage". (Those in know have seen the pattern multiple times.)<p>The answer is simple: find a similar field and <i>accept</i> a learning curve for the employee to transition. If you need something quick, hire a consultant to both move the project along and to mentor the new employee.
The real reason is that coding is hard and coders are hard to find and certifications are just an acknowledgment that you have <i>domain knowledge</i>.<p>I don’t care if you know what an auto scaling group is. I care that you can write code that automates builds and deployments. The AWS part is easy and you can learn what you need to know on the job. In my experience, coding is something you either can do or you can’t, and if you come to me with a cert and no programming experience it’s very hard for me to know what you’re capable of.<p>i spend like 5% of an interview asking about AWS stuff and like 50% talking about code and automation.
“server admin is very hard, and mostly gone with virtualization.“
I would say this is simply not true. For many companies, getting good admins and infrastructure experts this is extremely hard, at least in big corporate.
I'm guessing that another reason is that some of the existing IT personnel are resisting the change, and one of their methods is to give HR unrealistic requirements so no one will ever qualify.
Not so sure Docker is the only way forward when it comes to cloud scaling or deployment. VM templates are a very good alternative. They are more stable, more flexible/customizable and integrate more smoothly with CI.
I am sick of this non sense. It's because nobody knows what they want and those that should be the ones knowing it all also do not know what they know and what they need to learn.<p>We built immensely complex systems that allow us to scale. The thing is we have increased the complexity so much that we no longer can justify hiring incompetent.<p>Simplicity is beautiful and we will have a renaissance of simpler systems. Something along the lines of what Go brought to the game.