> the employers don’t ask “what you are able to do?”, instead of “what are your credentials/studies”.
I believe this issue extends beyond the scope of those without CS degrees; it also affects developers who have spent time working at small startups that haven't significantly scaled. If you haven't had the opportunity to work on large-scale projects, you might lack impressive accomplishments or credentials to showcase. This issue can compound over time, leaving you stuck in positions that are less rewarding than even mediocre jobs, regardless of your true capabilities or the quality of your work.
Some people say that you can build open source projects or build products to show what you have. Balancing such efforts with a full-time job can be challenging, as these undertakings can consume most, if not all, of your free time. This creates a vicious cycle where you're continuously overextending yourself just to prove your worth. It's a demanding situation that can impact both your professional and personal life.
Referring to the "small startup experience" not mattering - I would say it is almost like an alternate track as an engineer.
I have trouble landing interviews at FAANGs with 15+ years as an eng. and successful acquisitions as a founder - but those don't mean anything if you don't have direct knowledge of certain tools to handle scale or experience guiding projects across departments, etc. Which makes sense for any role past junior, since part of the value you bring is experience and only some of your experience is directly applicable. Vaguely being competent doesn't really move the needle.
On the other hand, I've heard stories of people getting hired from a FAANG going to a startup and not able to shift their mindset to a "make it work" approach, so the challenge goes both ways.
I've also had the experience of failing FAANG interviews after being very successful at smaller scale. Since then I've made the leap, and I can tell you that experience in the 10-100 developer range is extremely valuable and can make you very upwardly mobile once you get a foot in the door. People that have spent their entire career in big corporations often have a lot of assumed constraints based on groupthink and the vagaries of whatever specific org leaf nodes they've been exposed to. Often they will have worked for years under bosses with no real clue about the actual decision making process and strategy behind their work. It's extremely hard to develop an end-to-end understanding and strategic viewpoint in these environments. Startups offer much more opportunity to learn the big picture and the strategic considerations behind different functions. Obviously anyone can claim to be doing a startup—there's a lot of the blind leading the blind, and people playing house—but if you find one with an interesting vision, a bit of traction, and good colleagues you can learn lessons that will translate well to leadership at big tech.
Coming in the front door will be hard though, because recruiters and the first layer of technical interviewers will likely not have context to judge your 15+ years of experience. Look for referrals and directly talking to hiring managers. Also, read some writing from notable SV management writers like Will Larson and Camille Fournier, this will help you learn some of the shibboleths and how hiring managers think. Finally, if you don't have legit scale-up experience, look for that first. There's a range of companies with engineering teams of all sizes which can provide good stepping stones for learning as well as hireability optics.
> On the other hand, I've heard stories of people getting hired from a FAANG going to a startup and not able to shift their mindset to a "make it work" approach, so the challenge goes both ways.
But at least they get the job, so they can try to learn. Hence it doesn't go both ways.
> Referring to the "small startup experience" not mattering - I would say it is almost like an alternate track as an engineer.
It can be but doesn't have to be. A lot of people prefer to hire folks with some small startup experience because such people almost always better at thinking about businesses holistically. Many people who spend their entire careers at big companies can only reason about their little slice (lots of bad takes on businesses here).
A common way to make VP of engineering at big tech is the following 5 job hops:
1. Manager at big tech
2. Manager at startup -> promotion to director
3. Director at big tech
4. Director at startup -> promotion to VP
5. VP at big tech
The problem comes when someone gets too senior with relevant experience at scale. I (generally) can't hire someone with 15 years of experience who's never worked at scale. They're going to be bored out of their mind with mid-level work but simultaneously I can't actually trust them to not screw up something in senior+ roles, either technically or socially.
Titles in startups really don't map meaningfully to titles in BigTech. You can be called a VP in a startup with 20 people under you in the org chart. That empire size is equivalent to "second level manager" in a FAANG, and FAANGs know this. When FAANG acquires a startup with 20 people, MAYBE the CEO gets a nice title, but most of the startup's acquired employees get Senior Worker Bee titles.
I'd hoped the pre-requisite that titles matched duties was implicitly understood in the message, sorry!
But you can absolutely come into an engineering role at a company with 40 engineers (40 person org charts to director at BigTech), grow into leadership of a 200+ person engineering (charts to vp) and then go back to BigTech for bigger bucks.
It's probably like a lot of other transitions to related but different careers. In my experience, sometimes it works and sometimes it doesn't--which probably says that as a prospective employer (and employee) there is an additional risk factor.
Probably an unpopular, "grass is always greener" opinion, but this is why leetcode-style interviews can be a good thing at FAANG. Anyone with time and an internet connection can gain experience with data structures and algorithms via leetcode, whereas it's much harder to get hands-on experience with scaling real transactions-per-second on your own. And, as contrived as leetcode is, it does measure something about your ability to grok how code and infra scale against time and space.
The people I know (US-based developers) hired by FAANG were all very recently out of college, graduate school, or academia. Everyone else is focusing their time on diving deep into the technical issues they are faced with...
that and most FOSS projects are also not large-scale, or as a contributor you are only working on small patches that hardly qualify as large-scale experience.
Or they don’t see the scale, which leads to friction with users. You don’t see how your tool gets used, leading to DevEx issues with the project, people asking with varying levels of politeness why you built things that way.
But at the end of the day the feedback is only so valuable, even when it comes with PRs.
> If you haven't had the opportunity to work on large-scale projects, you might lack impressive accomplishments or credentials to showcase.
Whats a large scale project and why should they be viewed as an accomplishment?
I myself have pioneered the use of some tech for use in situations where it was not thought possible, I've worked on big projects that have included the largest in the world, for employers that include being listed in the stock market.
>I believe this issue extends beyond the scope of those without CS degrees
Does that make me worth a $718,000 Google Software Engineer or a $100,000 junior coder?
What price do you place on exploiting the vagaries of law that results in new legislation being created to curtail your activities?
The founders of Google, Facebook, Microsoft & Apple all started with no university degree, but they did all end up with stock market listings and some legislation to curtail their activities.
So do University degrees limit your earning potential whilst making it easier to find a job, whilst having no university degree seems to increase your chances of getting a stock market listing and some legislation to curtail your activities.
The founder of Microsoft, has wealthy connected parents and grandparents.
The founder of Facebook too
The founder of Google grew up in Academic family.
--
All of them received financial support, reducing risks loosing living expenses.
Risk of loosing livelyhood, significanly changes how a founder or an enegineer spends their time.
Connecting young people to networks of rich, influencial and generally experienced & smart people -- is a huge deal as well.
It improves their verbal skills, perception management, broadness their knowlege.
All of the above, can probably be easily found in many Stanford, Harvard, MIT graduates...
None of the are 'competitively obtained skills' due to 'hard work'.
--
Additionally moral choices that favor lying, fudging facts and intentions -- may also be a characterstics of graduates from those types of schools... This an opinion.
---
All of the above contributes to successfull financial outcomes.
None are particularly representative of person's merits compared to others...
Well... I recognise that these companies have made a lot of money, but how? I remember the day of BeOS, it was technically 15 years ahead and Microsoft killed it (just as an example), the same company that years later tried to kill Linux and killed Nokia. Google with their motto "Don't be evil" ... sure, just collect personal data in every possible way and sell it against your will. Yes, they have also done many positive things like Golang or Android for example, but when they see that they are losing the grip (and monopoly) on the project, they switch to something else (Flutter ... Fuchsia).
The real revolution (apart from the Internet) of the last decades has been open source, without it we wouldn't have *BSD, Mariadb, PostgresQL, Linux, Android, dozens of languages, databases, knowledge sharing. and with it freedom : freedom to choose, freedom to learn, to collaborate with other people around the globe for the benifits if not all, at least many. In my opinion, the open source philosophy has done more for human evolution than all the big companies put together (and their money stacked in foreing countries to not pay taxes).
My understanding, but maybe its fake news over here in the UK, is that Jobs dropped out, Zuckerberg dropped out, Gates dropped out, Brin and Page both dropped out. Cant comment on the other two MS blokes, or the other bloke who co founded Apple.
But getting a decent education here in the UK, where you dont get beaten up by teachers, in my experience is quite difficult. I even dropped out of one schools six form to enroll in another six form, perhaps demonstrating my willingness to learn, but even one of those courses was sabotaged.
To be honest I think the mentality of the british empire was and might still be pervasive with state school education, viewed more as an institution to contain people of a certain age than a place of learning.
Perhaps that is why law is not taught to everyone at school even in a TL;DR fashion?
My larger point is there is a very big difference between dropping out of an elite institution just before the finish line because you think you have a billion dollar company on your hands, and just dropping out. Even if MS, Facebook and Google all failed Brinn, Zuckerberg, and Gates would still have been able to walk get a job probably. They didn't drop out because they failed out, they willingly left to pursue their ideas, and were otherwise top notch students.
Not just that. Engineers get pigeonholed into tech stacks as well. It's almost as if companies (recruiters, really) think a senior engineer can't learn another programming language or something. They come at you with "we're looking for someone with strong golang experience," when there's no way to get "strong golang experience" without working someplace that uses it.
And? There's a learning curve any time one starts a new job with a new code base. Do you think a senior engineer can't become productive in a new codebase with a new language in a couple of months? If so, I'm going to guess you've never worked with a good senior engineer.
I can resonate with that. I regret not pushing harder to get into the bigger companies earlier. The scope increase in invaluable. You are not going to get it by urself unless u r extremely lucky and be able to scale ur own business to the comparable degree
Are you actually asking a question relevant to the thread or just spamming a self-promotion?
I ask because I don't see any reason that paid open source contributions wouldn't be as good as non-paid open source contributions when it comes to gaining/demonstrating certain types of dev experience (though ofc without the benefit of also potentially showing character traits related to being willing to donate time for free), so I'm struggling to imagine you wrote that comment actually hoping for any answer other than "great thanks for telling us about the thing you founded". Apologies if I've incorrectly jumped to a bad faith interpretation - if so, you could maybe be more clear in explaining what your question actually is; if my assumption was correct then I'd suggest thinking twice before using low-effort and low-value comments for promotion, since people (and in particular people who would be valuable to have using your service) are liable to be put off by any product/company/service that spams in such a way.
I found the comment valuable. I might be unmotivated to contribute to an OS project purely for the sake of doing so. But throwing some money in, I might go like "Alright I'll tackle this".
Certainly I'm at least glad a tool like that exists.
> If you haven't had the opportunity to work on large-scale projects...
open source projects/companies that pay contributors are oftentimes larger-scale
> Balancing such efforts with a full-time job can be challenging, as these undertakings can consume most, if not all, of your free time...
i'm asking whether some compensation could make contributing to open source less of a luxury for a lot of people (those with jobs but also those in the jobs market) and more of an accessible opportunity to publicly showcase your "true capabilities or the quality of your work"
FWIW, your comment didn't seem like self-promo or spamming to me. In fact, I consider it to be on-topic because it plugs an important gap that has always existed for people that intend to contribute to OSS but cannot due to their immediate circumstances.
The "bug bounty" model is already used for eliciting security contributions from security researchers so it is a good thing you've extended it to elicit general contributions from devs looking to contribute to OSS and be paid for it.
BTW, there's an over-representation of TypeScript and Rust amongst the projects you have listed. I'm curious is this because you already have strong network in these communities, hence the over-representation or it is a consequence of something else?
I track a lot of COSS projects and I wouldn't say most of them use those languages.
For interactivity, JavaScript is almost unavoidable on the frontend so it is understandable that teams use TypeScript (that gets transpiled into JS) on the frontend, but I would be hesitant to say most projects use TypeScript or Rust on the backend.
I know of several COSS projects that are written in Java, Python, Golang, C++. These are language that are not yet represented by Algora so there is plenty of room for you (or a competitor) to grow the pie of paid contributions to the COSS landscape.
If you have a dusty old blog with just a single technical post on it from 2012 which demonstrates that you know how to code, know how to write and know how to solve a technical problem... that will STILL give you a solid advantage against other candidates, at least in the early stages of the hiring process.
When hiring, the very first filter you need to apply is "can this person actually do the work". That's what FizzBuzz is about!
A blog post (or a dusty old GitHub repo) is a great way to jump past that filter.
I’ve been blogging for almost 20 years. The site isn’t monetized, I don’t have many readers, and I don’t get much Google traffic. But ...
- My posts about developing for iOS in 2008 were found by an editor at Manning and I wrote a book for them. This helped me make the transition to full time iOS development.
- My posts about extending Fogbugz and Citydesk were part of my application to Fogcreek/Trello — I can’t say it “got” me the job, but I’m sure helped me get an interview.
- I’ve been invited to speak at conferences, be a guest on podcasts, and write for Smashing based on people finding my site. It’s not a lot, but raises my profile just a bit.
I would do it anyway, but I’ve had a 30+ year career and I think my blog has been a significant factor in it.
I wrote a ton as an industry analyst as well as lots of articles/blogs. Absolutely was probably the factor in my getting a long-time job. Maybe indirectly but it's why I knew people at the company and they knew me.
Well ... I think we have many problems in these modern IT times ... The first is that we are drowning in new IT technologies that are released every day and it is impossible to keep up with them ... Many languages, frameworks, tools ... After 25 years in this field, I am very sceptical when I read a CV on LinkedIn of a person who is an "expert" in 10 or 20 different technologies or languages (it is simply impossible). On the other hand, we have companies looking for "gurus" in several areas with many years of experience... just unrealistic expectations.
Apart from a limited minority of true geniuses, all of us mere mortals need time and projects and work and a lot of failures to develop expertise in any field.
On the other hand, in the 'real' world of work, it is not enough just to be technically competent. You need a lot of other skills that are equally if not more important: willingness to learn, ability to evaluate technologies based on requirements and constraints, integrity, honesty, self-organisation, curiosity, ingenuity ... logic. At the end of the day, it is a balancing act between technology, resources, money, time constraints and human relations, not just writing code or knowing everything about one or more tools that may become obsolete in a week.
It does feel kinda overwhelming. You really need to run to stay in place, and it is more and more about setting up a brittle chain of config files than having deep knowledge of a single languages best practices.
Well, the worst thing in my opinion is that (often for ideological reasons or egocentric positions) we don't accept or don't want to improve the existing tools or languages, but we create a new one ... so we had C, C++, Object Pascal and now we also have Python, PHP, Ruby, Go, Rust, Nim, Zig, D, Objective C, C#, Java, Kotlin, Haskell, Erlang, Ada, to name a few ... and this is the easy side ... Now there is the bandwagon of javascript frameworks: React, Vue, Svelte, Next, Nuxt and so on. It is impossible to keep up and it is also impossible that all these new technologies will be maintained in the long term. I remember the rise of Java years ago ... Java everywhere: Java applets on the web, Java on the desktop, Java as an application server and how it ended? Java is, in my opinion, one of the most convoluted and impenetrable languages ever created, but when you work alone as a freelancer, as in my case, you don't have the time and the resources to embrace and follow every new technology that is marketed as a revolution and the future, because at the end of the month you have to send the invoices to your clients to earn and live, and you have to have produced some results that work for them in the "real" world.
I have used Drupal for a few projects in the past: the main problem? Every major release that broke the API backwards compatibility and this means you have to refactor all your plugins and custom code, many times without any new direct benefit to the end user who, of course, was not available to spend money because some internals change every year and, for him, on the surface nothing changed.
You realize that Ada and Objective C predate Object Pascal, and Erlang came out in the same year as it, right?
I would worry less about new ways of doing the same thing and more about client generation/retention. Plenty of companies are willing to pay top dollar for boring technology.
It wasn't a chronological order, of course...
But if before the languages were more or less field specific, now with each language you can build any kind of application and it became very difficult to make a choice or simply keep up with the daily updates.
Congratulations on the job. I had a similar experience where having a public website with published extensions for the server that my future employer was using helped me get a job there, but, I'd like to point out that in both our cases it wasn't the blog/website alone that helped us get the job, but a personal connection. In your case it was meeting that CEO on a bike, who then made a personal recommendation. In my case it was my general reputation as an active member of a community where the job was posted, that I applied to.
What the blog did was to make it easy for the CEO to get to know you without talking to you more. And for me it was my code examples. So that works if you find a job where your ideas fit, but it can also work against you to find any other job that may expect a different mindset. Just recently I had an interview that also came about due to my reputation in the community, after I posted there that I was looking for work. But they desperately wanted to see code too, and the only thing I had to show was almost 15 years old or older, in a language that is not well known, because all my recent work was not public, and I didn't keep code from work on my personal devices. I haven't heard back from them yet.
Minor feedback that you're welcome to ignore: I found reading your comment a little bit slower & more annoying because of your lack of upper case letters for sentence starts. It's an interesting comment so I still read to the end, and upvoted, but I was close to not bothering to keep going half way through purely because of it all being lower case. I'd suggest changing your habit and starting sentences with upper case letters, not just because it's "normal" but because it's normal for a reason, that reason being that people find it easier to read.
I appreciate the feedback. This is a habit that I acquired decades ago, on the basis that it doesn't make sense to write words differently just because of their position or their meaning when there is no such difference in the spoken language (but there are pauses between sentences instead, which I didn't consider then). It is actually through reading HN that I started to consider readability of longer paragraphs. I am considering the situation. Don't know where I will end up.
I edited the comment, but now I am stuck between leaving the commented with capitalization to improve readability and reverting the edit in order for your response to make sense and to allow others to judge for themselves. (I'll go with keeping the edits. Anyone curious about the readability can go look for older comments of mine)
I think this is how I imagine hiring to be but it is not scalable. You get hired for doing things you love and have passion for. The fire in the belly comes from passion and love for the craft.
Lot of engineers I have interviewed, if feels to me they are trying to hack the interview process. In larger orgs, "cracking" the interview process might lead to long term job. But in a mid tier or start up, that's the beginning.
Yes, I totally agree with you, this method is good for small/medium sized company but here in Italy the majority of companies are small/medium, so I think that hiring based on the “real knowledge” should become a prevalent strategy to growth up. Instead we -unfortunately- have more recruitments based on the “I’m a friend of a friend” or “this piece of paper says that I’m able to do to xyz, no need to prove it”.
Degrees most certainly prove that one is able to do a significant amount of boring tasks. They also prove a being beyond minimum intelligence and skill level. Safe choice but often not the best.
Hey, about the raspberry pi, I have had a bunch of problems with it eating the micro SD cards, like they fail in six months or an year tops. I had countered this for some little time by using it to boot and relaying to a hdd. How are you handling this? Is there some special micro SD card that doesn't fail as soon?
* SLC - 1 bit per electrical cell
* MLC - 2 bits per cell
* TLC - 3 bits per cell
* QLC - 4 bits per cell
You need SLC cards if you really want something to be long-lasting. You can find some at specialty places like Digikey, or labeled as "Industrial" in normal channels (i.e. Amazon sells both Kingston and Samsung 8GB industrial cards).
They are both smaller capacity and more expensive, though.
Also, if you're using a stock distribution not intended for a Pi, you probably have quite a lot of logs/etc churning, which can be cleaned up. If you aren't doing much with the Pi except a few common services, check out DietPi.
Thank you so much, I didn't knew about this! Also on DietPi it's pretty cool and apparently it has also builds for x86 so I can easily play with it in a VM for experimenting! Thanks for this!
Standard flash disks use wear leveling - which is not normally present on micro SD cards. Have a look at the SD cards that are designed for dash cams, body cams and industrial applications
SanDisk used FTL (flash translation layer, wear leveling) on their SD cards at least way back in 2013. Single MLC block survived about 500 erases. FAT would kill it immediately otherwise.
I went down that rabbit hole a few years ago. Many cards don't tolerate sudden power loss while writes are happening.
The solution I found was these ATP micro SD cards (looks like the prices have increased like 3x though...) where the firmware is specifically designed to tolerate abuse:
From what I've read, the power supply can make a big difference here, in addition to getting quality microSD cards.
I've had a Pi Hole running on the same microSD for 6-7 years now with only one hiccup (and that was a software thing - a fresh install got it back up and running on the same microSD). It just has a regular SD card (not a high endurance one), but the power supply came from adafruit, and it's been connected to a UPS for most of its life.
The good power supply is a good thing, I used old Samsung USB chargers, I imagine they deliver an ok quality, but perhaps not as much as a phone has a big battery that can work as "big capacitor"/filter of sorts. The UPS is a good idea too, I probably need one here and power going out where I live is something that does happen occasionally.
That's a good thing to check. I honestly don't know if the SanDisk micro SD cards I got are genuine - they look their part and have that plastic square thing that is placed inside the paper thing with SanDisk brand and all the numbers, but maybe the fake ones have caught up in their packaging. I have no idea how to make sure it's genuine though.
Congrats on the new job, I’ve read the blog and it is truly well deseeved.
Regarding on the difficulties of the italian market, I’d like to dissent though.
Fact is, in Italy is waaaaay more simple to cheat your way to the top instead of having to go through the pains of getting a proper degree.
I know because I did it and went from a non graduated outbound call center operator to dirigential positions in highly respected fashionable industries in less than 10 years.
Point is, and with the exception of research roles o and a fistful of high consultancies, who enforce the control on graduation, most likely than not you will not be asked for proof of your studies if your jobtitle carries enough weight and if your accomplishments a friendship with the right people weights enough.
Trick is to change joq with the right frequency in the first years, participate to associations and try to be in the board, blog a lot in the right environments (and specifically there), make a lot of acquaintances and help them solve tricky situations for free, and generally speaking be as visible as possible as a tech savyy and strategist.
At the right moment, the reputation you will have accrued will be enough to receive accreditations beyond the first line of HR, and id your sponsor is notable enough no one will ask for your degree.
Once you have done that the first time, is the weight of your jobtitle and your name that propels you forward, no one willing to question your perceived authority.
Than rinse and repeat until the weight of jobtitle and name combined is unquestionable.
If you know how to behave, you will rise fast.
And good luck to trying that abroad. On a sidenote, once you earned the jobtitle this way in italy, it is easier (but still tricky) to get it abroad as well.
Once I tried to go abroad sometimes I got asked credentials (and subsequently bombed) and sometimes not, wich guaranteed me jobs in Switzerland and France.
Steer away from anglosaxon jobs though, they do enforce the control and most likely they’ll carch your bluff.
last suggestion: shoot for the highest possible pay and change job often enough to beat inflation several times over in 5 years. Above 150k/years almost no one will ask you for degrees. Only your pay and jobtitle and experience will count.
> the employers don’t ask “what you are able to do?”, instead of “what are your credentials/studies”.
I think this is cultural.
Italian culture is hierarchical and trust is built upon relationships, you either need the authority of a good cv or a good network to reap opportunities. (This is quite common in southern europe)
Northern europeans, americans etc are more egalitarian and have a task based approach to build trust, and value ones worth accordingly.
I think this is universal. I read the story in a different light: it's all about your social graph and who you know.
This is where I see most of the value from college or school, building up your social graph of people in your approximate industry.
Want to be successful? Connect in a meaningful way with more people. Your opportunities will multiply.
Don't know how to do that? There's an easy shortcut: publish something! Give a talk, write a blog, teach a class. Find those opportunities where you can engage with many folks at once (not just 1:1, that doesn't scale as fast).
It depends on what motivates people from your network to give you an opportunity.
If their motivation is “I know that this guy is good for sure, I don't have to take chances with random people who answer the job as”, it is still merit based, and they're acting in the best interest of the company. However, they could also give a job to somebody they know just to have their favor, or to get more people into the company they they could rely on in internal politics. In this case, this is not based on merit, and they're effectively defrauding the company, acting against its interest.
Personally, I've mostly encountered first type of networking in USA and Northern Europe and mostly second type elsewhere. Israel is weirdly split between Hi-Tech and everything else, following first and second type respectively.
It's been over ten years now since I blogged recently, alas, but back in the day blogging got me a book deal and being asked to chair an O'Reilly conference. I recommend everyone blog about what they're into because it's both personally edifying, but it also only takes the right person reading to reach out with an interesting opportunity to be more than worth it.
Congratulations. Your articles are always a pleasure to read. Reading about your homelab really changes from the "Here is my 400 cores homelab" that often appear on HN.
Ahah thanks, yeah if you don’t need 400 cores, why have them? I mean, sometimes you need power but if you don’t need it, is better to have a slightly older hardware, but more reliable and, above all, with less consumptions/environment-bills impact.
In example: I’m using at home a RaspberryPI 4B as a Pi-Hole that is a totally waste of power, I’ve installed the same sistem in some companies with 70/80 DNS clients and all is running more than fine …but on a Raspberry PI 3B!
Congrats! I hope everyone can see he actually found a job due to getting out and meeting people (in this case, he met a CEO while they were both cycling), who then became interested in his work due to the relationship he fostered.
Thanks! Really glad that you read and appreciate my posts, unfortunately I don't have always much time to write but when I can, and I have something to share, I try to write and post useful info.
Yesterday you are shitting on someone's home networking and today you learn he made it his job.
Regarding the credentials... you would always find people who are dumb^W think what the certs are the measure of quality. Sometimes it's actually good, if a person asks for certs first then you know you don't want to work with that person.
NB: never send alerts directly to people.
Create a group (distribution list, alias, whatever it's called in your mail server), preferably multiple groups foreach client, eg: pasta.zara.alerts.monitoring@giuliomagnifico.com etc
That way you would see which instance generated the alert even if you didn't configured the custom alert message,
you can filter all these messages (to tag, move to folders or whatever) simply by filtering on *alerts.monitoring@,
and if for some reason you depart with the client in not so amicable ways then you can simply kill all the mails from it by a rule with $To:.
Regarding the hardware, I would not recommend to use a raspberry pi, not only are they hard to get and expensive but as another comment states the RPi rely on SD cards which are prone to corruption. I think it's better to rely on second hand enterprise hardware if you're looking for something cheap.
FWIW, Raspberry Pi availability is starting to recover, and they do have the ability to boot off of a USB drive (including SSDs) or a network server these days.
That said even SD cards are probably fine if you're not doing a ton of writes. (And it looked like a USB drive was attached to one of the PIs in the picture. So, presumably the important data is going there.)
You're not wrong that second hand enterprise hardware would be more reliable but, honestly, I think this setup is fine for what they were trying to accomplish - especially if they already had some of the hardware laying around.
I have used 4 RaspberryPi (3 and 4) for a few years for various needs : Gitea server, Home Assistant server, Syncthing main repository, Borg backup repository, Wireguard VPN and network gateway between 2 LANs. I haven't encountered any problems with the SD card, if you buy the class 10 SD cards (the same you would use for your dashcam and security camera).
One option is to use the F2FS file system on the card (I did this on my Raspberry) or, as others have suggested, use an SSD or NVMe drive. The main problems I encountered with the Raspberry 4 were heat dissipation (use a metal case with direct contact to the CPU without a fan, this is the best option in my tests) and external USB 3 drives interfering with the wifi. The latter seems to be a known design issue.
Considering all this and the prices, if you don't really need the Raspberry's ability to control external hardware via GPio, and if you want to keep the same network design (separate devices), there are many mini PCs with passive dissipation, 2 or more Nics, lots of RAM and SSD support, for a similar or lower price all considered ... or buy one or two real servers and virtualise everything with Proxmox.
Now they aren't expensive (like 6-12 months ago), one RPi -the one with the database- is relying on the SSD (so no issue) and the other is using RAM logs so the data written to the SD card are minimal.
I'm running a Raspberry PI with an SD card and RAM logs since 6-7 years with the same SD without any failure.
this is advice I hear very often, but where do you buy second hand enterprise hardware?
I went and bought mine second hand from the internet but ive got no idea where to get enterprise stuff
I got an old Dell PowerEdge 610 server second hand on gumtree.com.au (classified ads page, I guess like craigslist or something?) for $400. The guy selling it had bought it for his home theater without realising that commercial servers are (a) noisy as hell, and (b) useless for graphics. :P It's been running our half-dozen server VMs at work for the past 3 years and it still blows my mind that you can get a server with 12 CPU cores and 96GB of RAM for that price.
Edit: 610, not 670, sorry. It basically sits in a corner making a faint whirring sound and apart from occasional apt-get upgrades I tend to forget it exists.
If you're in a major city, look for salvage outlets. There are usually operations who recover decommissioned hardware from businesses, and then thrift-store style, sell it to hobbyists.
I see power input going into the Mikrotik router, then feeding D-Link USB hub with (according to the Mikrotik specs) up to 5V/500mA DC.
The USB hub sends power to both Pi’s and somehow splits the 500mA between RPi 3 (officially needing up to 5V/2.5A) and RPi 4(5V/3A). And it still somehow works?
I know all the specs are not set it stone. USB can feed more than 500mA and Pi’s can run on well under their requirement but I’m no stranger to Pi’s and I had experienced power issues due to inadequate supply before. Especially when the CPU is maxed out, like during the boot sequence.
Am I missing something here? This looks quite insane.
The 5V/500ma number is copied from the USB 2.0 spec; the Mikrotik device may be outputting more than specified by coincidence of the hardware used to manufactured it
I agree but how do you explain that it’s actually working? Also on heavy load for days? We was asking the same, for that I tried to run an heavy benchmark for some days on the RPis. Maybe the spec of the power supply are different?
I was wondering the same but then I discovered that the Mikrotik is able to power both the Raspberry PIs when in full load for 3 days straight. I did some benchmarks, but I think the power output of the Mikrotik is different, more than the official specs.
Grazie mille =] I hope to find the time to blogging! I must found it, sharing knowledge is a very important thing to help each others for a (little) better internet.
i live in a more northen country, but still writing a small blog helped me too to get my first IT job. a friend gave me the tip for me to show i had recent relevant,activities for a job. its a really good tip. even if no one reads it you can put it on your resume just as a proof of your words. best tip i ever had. happy it helped you in this way and thanks for sharing!
Congrats. I have a few questions though. How do you plan to transition from hobby and consumer grade products to business grade products with this setup? With 24/7 operation, those hardware will die within a few months. Do you plan to make a second redundancy box just in case?
Congrats Giulio!
Incredibly, things are changing even in Italy.
I find saddening that the majority of the youth is either stuck in dead end jobs or in lowly-paying slightly career-related jobs.
I try to push anyone I meet under 30 to just leave the country. The only other option is to embrace entrepreneurship, even though it requires one to grow a thick skin, stomach and an indestructible carapace.
I might be wrong trying to de-populate a country of old farts, but everyone has the right to strive for the life they want to lead, or at least get to choose a job/career.
One last note of folklore - I find Italian politics amusing: both left and right have been focusing on civil rights debate, without a grand scheme but rather a day-to-day routine of putting out fires as they come along (the motto being 'just kick the can down the road').
We need leaders with a vision, not social media influencers that follow behind whatever is trending on social media this week (aka the flavour of the week).
Once again, congratulations! :)
Sorry for hijacking the celebrations with a rant.
Sure. I am asking, whether we could get corrected titles here on HN for such cases.
I personally think, that a little due diligence is appropriate for titles of such blog posts, but that is only me and what I would do, if I published a blog post like that.
The author lives in Italy. From the first paragraph:
This maybe seems to be a weird post, especially for those who don’t live in Italy, because here in Italy there’s the common belief that the abilites of a worker, usually are less considered than the qualifies that it has. Or simply the employers don’t ask “what you are able to do?”, instead of “what are your credentials/studies”.
Thanks, it's what I wanted to achieve, so I'm glad I've reached it! By the way the blog is powered by Jekyll and theme is 'Jekyll Yat Theme' (https://github.com/jeffreytse/jekyll-theme-yat/) slightly customized.
I believe this issue extends beyond the scope of those without CS degrees; it also affects developers who have spent time working at small startups that haven't significantly scaled. If you haven't had the opportunity to work on large-scale projects, you might lack impressive accomplishments or credentials to showcase. This issue can compound over time, leaving you stuck in positions that are less rewarding than even mediocre jobs, regardless of your true capabilities or the quality of your work.
Some people say that you can build open source projects or build products to show what you have. Balancing such efforts with a full-time job can be challenging, as these undertakings can consume most, if not all, of your free time. This creates a vicious cycle where you're continuously overextending yourself just to prove your worth. It's a demanding situation that can impact both your professional and personal life.