Note: not all companies do this, most companies I meet do. Maybe I'm meeting the wrong companies.

Here's my original comment:

Well, the flip side is. Do you have experience with AWS? Not really, but I've known Linux for years, have done stuff on Heroku, Digital Ocean, my own servers, hacked boxes and IBM Bluemix, but no AWS.

Sorry, we rejected you because you have no experience with AWS.

So to be honest, I've been looking into it, because apparently a lot of Dutch (not all) are too short-sighted to see that I've done development and am able to figure it out, or have transferable knowledge (fun fact/example: in part via HN articles that showed how to load balancers [0]).

Also on the money side, it's not that bad. It's $150 for an exam. I think by reading docs and watching this video [1]. It'll be possible to get.

Everything I mentioned here is also true for pentesting as I've experienced something similar there.

Now that you've heard me complain, I am actually for hire ;-)

I'm language agnostic, most experience with NodeJS/ReactJS, junior pentesting as second skill, I'm at my best as a 50% programmer / 50% anything else <-- pentester may be included in the second 50%, or: marketing, business analyst, UX, I'm open-minded and willing to learn.

If no company is hiring me by the first of January and all my applications are dead in the water, I'm starting something for myself.

[0] https://news.ycombinator.com/item?id=21490731

[1] https://www.youtube.com/watch?v=RrKRN9zRBWs&t=32838s&ab_chan...

I think this is a great point and the economics of the job come into play too...

A lot of the initial challenge in interviewing is proving you can meet a certain bar to get your foot in the door for an interview. And $150 to prove that you can do a 100K+ /year job seems like a pretty rational investment

I'm from NL, so it's at best 50K. Though with the current USD/EUR, I guess that's not bad at all.

I don't doubt your technical skills, this comment is just to help you see the side of the employer. They don't have time to onboard someone on AWS, they need someone to be able to speak "AWS" out of the gate and contribute immediately. General knowledge isn't helpful, when you need to be able to fine tune each AWS specific services.

The problem is, one with transferable skills can get themselves to "speak AWS" in maybe a few hours, once they know exactly what they are going to use (or what needs to be done).

Those “skills” are only transferable if all you plan to do is spin up a few VMs. Have you looked at all of the different services that AWS as well Azure and GCP offers above just hosting services?

Anyone who comes to any cloud provider with that mentality actually is worse than someone who has no experience.

That’s just as bad as someone who models a schema for a database meant for analytics the same way that model a database meant for transactions and wonders why it isn’t performant.

> Have you looked at all of the different services that AWS as well Azure and GCP offers above just hosting services?

I have checked their huge product offerings a few times when I needed something I know they must have, and used that, like AWS Batch and AWS SQS.

As for "all of the different services". No, I haven't, but I can, and I will do that, on demand. My point is exactly that I don't want to sink huge time into reading their documentations for something I may never use. And I don't believe it is correct to prefer candidates who did that.

The problem isn’t knowing every service in detail. It is knowing the “what” and not necessarily the “how”.

But the attitude that a cloud provider is just a bunch of VMs leads to crap like hosting everything on VMs, reinventing the wheel etc.

You wouldn’t hire a database admin with no database experience and expect them to learn on the fly. Why hire someone with no experience with your cloud provider and hope they would learn on the fly?

Any technology/service/language has footguns and best practices that you only learn from experience.

I think you and the comment above lay out both sides of the argument pretty succinctly.

AWS is more than a bunch of servers. AWS has over 160 services. Saying that you have experience with Heroku and DO as an argument for why AWS shouldn’t be a problem is exactly the reason I wouldn’t hire someone without AWS experience.

What are you interested starting? Dutchie currently in australia saving to start my own thing (relocating cheap country), looking for a cofounder. Would love to discuss. riekus at the proton provider.

Dutchie here (currently in australia) also into WHM and meditation. Planning to startup in February, want to discuss? Riekus at proton mail.

To me it is akin to saying:

1921: "Why learn a Underwood typewriter and learn all their marketing junk?"

It just actually means you want to learn how to type. Sure the brand you learn on is a "Underwood" - but the concepts and structure are transferable. YES! You don't have a CAPS lock feature.

Cloud companies are just as weird and envious as the rest of us. So they copy!! If you can learn s3 you wont be shocked by any principals of gcs. Cloud tech is transferable. Cloud containers? Docker works on both.

Knowledge is transferable between languages too but it still entails ramp up time and employers will discriminate based on experience. For example knowing the nuances of PHP don’t necessarily prepare you for elixir all that well. Likewise, someone SSHing into their VPS and installing nginx isn’t necessarily experienced with k8s and IAM

Languages and paradigms are now almost entirely in a separate camp of comparison from AWS vs Azure. How did we get here? We kicked the sh*t out of commercial languages so we wouldn't be indentured servants to a small number of companies.

People tend to use binary logic to gloss over the details that determine our future.

K8s is opensource org backed, docker is opensource startup backed. sshing in and installing Apache is an open setup that can't be sabotaged but not open standard. Signing in to Linux and starting your own python or C++ server code is open standard and can't be completely hijacked even by a BDFL controlling the main implementation direction.

> For example, because of his bullshit aws certificates, this person will have the feeling of being part of a special club, and a sunk cost incentive to push for whatever possible aws solution to any problem instead of anything else.

I haven't met very many, but this is basically the attiude I've got from most people with AWS training.

Edit: With good/experienced developers, it's usually not that bad, but with new developers it can easily turn them into architecture astronauts.

We need to do it mostly because AWS requires it if you want to get into certain tiers with their partnership and/or publish images on the marketplace.

We rarely (if ever) actually communicate our AWS certifications, except towards AWS themselves.

This.

I got certs because the company wants them as part of their AWS contract. Other than that, the cert means nothing at my company.

Lots of companies offer training sessions on their products. I think it's reasonable that if you have a moderately complex product, you insist that your support team not waste time walking customers through elementary stuff they should have learned in a seminar at the outset.

Strongly agree with your sentiment. The number of jobs that require "expertise in AWS/expertise in Google Cloud/etc" is laughable. It's like every company is creating its own vocabulary for things network engineers have had terms for years.

This serves as an artificial barrier to hiring. I've done years of network engineering but then it comes down to "how many years of experience in AWS."

EDIT: Note that this isn't a new thing, e.g. we're looking for an "Oracle engineer"/"vmware specialist"/etc.

You realize there is a lot more to cloud providers than just networking? I had the privilege of working with old school netops people who took one certification, duplicated everything on AWS like they did on prem and ended up costing the company much more money in resources and not saving money by using other AWS services. Once I learned about all AWS had to offer besides what lift and shifters knew, I realized how much money there could be if I learned about the other services instead of thinking of AWS as just a bunch of VMs. The same could be said for Azure and GCP I just fell into AWS.

CCNE MCSD MCSE and the dozens of other corporate certifications.

It is very funny when aws came up with their own.

I think this happens when a company engineers something so complex you need to train people how to use it. Rephrased as when a company fails to make their products easily understandable.

Having said this there is a place for such complexity and with market share comes the right to do so.

The aws product line is quite comprehensive and they have broken new ground trying to offer this.

Of course if this was a google company I would imagine product creation and sunsets. Meaning they shouldn’t be doing this.

The Cisco stuff is annoying because when I try to get some solid networking knowledge most of the resources are Cisco product training materials.

Seems like a valid way to learn vendor's platform to me. If I was looking for an AWS person having credentials from Amazon would be a big selling feature.

I have co-worker who is retaking 5 or 6 AWS certification in 3 month. You may think it’s useless but the AWS discounts, credit and kickbacks we’re getting on our AWS projects makes it a no-brainer. The time and cost is much less that what we’re making from keeping our AWS partnership level.

Gaming the system... maybe, but it pure profit.

I disagree. Cert studying is a good way of learning and validating your knowledge of a certain providers tool stack and way of doing things. My company provided the learning time for me as a junior so that I could get up to speed with how they operate. I work with people who have a LOT more experience with Linux and enterprise networking, but they are understandably stumped if they have to deal with a cloud platform. It’s pretty confusing since everything has a ”name” and there are dependencies that aren’t obvious.

Calling it purely marketing content is cynical because even documentation or tech talks could be interpreted as marketing if you are cynical enough.

You can always choose not to go through their hoops and just study learning content but not book an exam. Learning platforms don’t really care.

Spot on. Certs are great at enforcing habits and validating what we know.

I look at it this way: I'm a noob. I learn the things I do because I've genuinely developed an interest and understanding of how technology can improve my life and that of others.

A Raspberry Pi home automation server running Alpine, a load-balanced OpenBSD router... I make some projects which are considered a hobby for now.

But how would you convey this to a recruiter with limited to no understanding of technology? How would it even look like to a recruiter if someone like me was trying to get a foot in the door?

This is where certs might come in handy. I may not become an expert overnight, but at least it shows that I'm willing to focus and learn the material required to pass a cert.

Your point is exactly what I'm saying: because of the network effect you are now required to get the useless certificates. And even if it is not your intention, you will also participate in the effect once you get it.

For example, if employers would not find any good candidate with the certificate, they would not think of it as valuable.

Also, on the general topic, what should be more valuable:

A) an aws certificate on Route53

B) a certicate of training in dns and dns configuration

Which one will give you transferable knowledge to the other? But which one will be favored by a clueless HR recruiter?

Well, take a look at the 160 services that AWS offers. How much would experience on your PI help?

I disagree. I've worked with two groups of professionals, those that have certification and those that say they "know" AWS or have working knowledge or experience. It's quite painful to work with the latter when you depend on their expertise.

Give the content a try, you'll find at the foundation of it is building highly scalable distributed system, just that the execution of it is via AWS. You can apply the same knowledge to any cloud platform.

> building highly scalable distributed system

I doubt that. Looking through the link and recently having discussed this in an OT-thread of an hobbyist-forum, the only thing you learn is AWS-lego. I don't see, how training for marketing keyword-riddled multiple choice-questions helps you in understanding the systems in question (case in point: Q1 of the "blog" - basically the only skill required is basic (!) logic matching requirements to the services (my mother can do this) and translating the marketing speech to requirements... – you don't need to know anything about interdependencies.)

At the end of the day, those that get hired because of their certificate can't actually take up budget resources and not be able to expertly talk about HA architecture and deliver on a technical roadmap. You can't hide behind a certificate. A technical professional actually needs to deliver a solution to a real-world business problem and when you don't it is highly visible. (Slow service, downtime, lost data, breaches)

If certification is indeed useless, and exams are simply basic marketing exams, it will eventually become a red flag in the hiring process.

> to learn all of a company marketing content

This may be true for the lower levels (haven't taken those). But if you go higher, it's pretty heavy technical content. On the second level, I'd estimate <10% of questions/content was marketing.

1) It's not that much money. 2) The skills are useful, unlike those in your flawed analogy. 3) Companies actually look at such things when recruiting contractors, etc. 4) You're an *.