Hey everyone,

During some backend improvements to the devRant infrastructure, some of our async queue processors (SQS) stopped working which caused many notifs to not go out/stop working. Unfortunately our alerting didn’t pick up on this since there were still queues being processed (just not specific ones) and some aspects of notifs working. Big apologies for this issue!

It is now resolved, and while very delayed, no notifications were lost and all were processed after the queue processors started up again. Sorry for the bulk notifs, but we wanted to make sure all that were supposed to go out went out.

Additional alerting will be put in place to prevent this from happening again.

Thanks for your patience!

So recently we re-orged to a product vs engineering (yes, I meant vs, it’s contentious) organizational structure. One of the former dev leads got picked for product and went on this lovely ass-kissing spiel about how great this was in front of our new bosses. The next day(!) he was telling his old team what to do directly to his buddy the scrum master, who works for me and casually mentioned it. How am I supposed to run engineering and deliver if every P.O. can end run around the structure? I hate all this.

Also, if the new PE tells me one more time all my problems can be solved with SQS, I’m gonna explode. Not all dev problems are a nail to fix with an sns hammer. Asynch comms has its uses, it is not the *only solution.

I feel like I’m over reacting, and yet, I still feel rage…and happy to find an anonymous place to rant about it.

Damn, how have I only just discovered localstack?

The ability to spin up and use SQS queues, S3 buckets, lambdas, Kinesis streams etc. for development without worrying about bankrupting myself if I screw something up is really quite liberating.

So recently I completed side gig from random freelancing site where I had to shadow troubleshoot performance problems over teams call with random Indian guy on his client's AWS account. Long story short you can autoscale new instances all you want but it's not gonna help if your FIFO sqs has only one message group ID. This architecture is running an online game, which is basically limited to processing ONE event at the same time for ALL players xD
What's even better, basing on naming convention I realized it's a company that I interviewed for like 4 months ago and they told me "we need someone with more experience". Well good luck, thanks for quick cash

AWS write awful docs. The .Net SQS example code has spelling mistakes, bits missing and doesn't work.

Also would be helpful if it highlighted that the .Net Core implementation is fundamentally different.

Jerks.

fuck u aws

all that money and u can't read idiot proof documentation for me to have SQS connect to a VPC endpoint to read a message

also fuck u search algorithms for not handing me an easy bake solution to what is a not a novel situation

also fuck me for being unable to git gud

Quick Tip:

There are jokes you tell to your best buddies that you should never tell your boss.

Just saying : |

Serverless and death of Programming?!

_TL;DR_
I hate serverless at work, love it at home, what's your advice?
- Is this the way things be from now on, suck it up.
- This will mature soon and Code will be king again.
- Look for legacy code work on big Java monolith or something.
- Do front-end which is not yet ruined.
- Start my own stuff.

_Long Rant_

Once one mechanic told me "I become mechanic to escape electrical engineering, but with modern cars...". I'm having similar feelings about programming now.

_Serverless Won_
All of the sudden everyone is doing Serverless, so I looked into it too, accidentally joined the company that does enterprise scale Serverless mostly.
First of all, I like serverless (AWS Lambda in specific) and what it enables - it makes 100% sense and 100% business sense for 80% of time.

So all is great? Not so much... I love it as independent developer, as it enables me to quickly launch products I would have been hesitant due to effort required before. However I hate it in my work - to be continued bellow...

_I'm fake engineer_
I love programming! I love writing code. I'm not really an engineer in the sense that I don't like hustle with tools and spending days fixing obscure environment issues, I rather strive for clean environment where there's nothing between me and code. Of course world is not perfect and I had to tolerate some amounts of hustle like Java and it's application servers, JVM issues, tools, environments... JS tools (although pain is not even close to Java), then it was Docker-ization abuse everywhere, but along the way it was more or less programming at the center. Code was the king, devOps and business skills become very important to developers but still second to code. Distinction here is not that I can't or don't do engineering, its that it requires effort, while coding is just natural thing that I can do with zero motivation.

_Programming is Dead?!_
Why I hate Serverless at work? Because it's a mess - I had a glimpse of this mess with microservices, but this is way worse...

On business/social level:
- First of all developers will be operations now and it's uphill battle to push for separation on business level and also infrastructure specifics are harder to isolate. I liked previous dev-devops collaboration before - everyone doing the thing that are better at.
- Devs now have to be good at code, devOps and business in many organisations.
- Shift of power balance - Code is no longer the king among developers and I'm seeing it now. Code quality drops, junior devs have too hard of the time to learn proper coding practices while AWS/Terraform/... is the main productivity factors. E.g. same code guru on code reviews in old days - respectable performer and source of Truth, now - rambling looser who couldn't get his lambda configured properly.

On not enjoying work:
- Lets start with fact - Code, Terraform, AWS, Business mess - you have to deal with all of it and with close to equal % amount of time now, I want to code mostly, at least 50% of time.
- Everything is in the air ("cloud computing" after all) - gone are the days of starting application and seeing results. Everything holds on assumptions that will only be tested in actual environment. Zero feedback loop - I assume I get this request/SQS message/..., I assume I have configured all the things correctly in sea of Terraform configs and modules from other repos - SQS queues, environment variables... I assume I taken in consideration tens of different terraform configurations of other lambdas/things that might be affected...

It's a such a pleasure now, after the work to open my code editor and work on my personal React.js app...

I'm taking a class in my university about Cloud computing. In 2 weeks we made a simple web app to upload videos and then a simple job that converts all videos to mp4.

Now we took the app to the Cloud using AWS. We created different instances for the web servers, we changed the database to NoSQL, used SQS to queue the convert videos jobs to the different workers instances, used SES, S3, CloudFront, ElastiCache. All that stuff.

And all that is worthless because I cannot get my Ubuntu instance to run a fucking command on reboot. I don't really know how and I feel that all my work was wasted.

Feels bad man

So i just learned aws elastic beanstalk (EBS, ECS, ALB, EC2, Amplify, S3, RDS, SQS)

Essentially i learned how to operate with aws to deploy a full stack web application with custom backend i built, with security and jwt token, certificate manager, ssl/tls to set up https and redirect from http, and react/angular/nextjs on frontend

All with custom CI/CD pipelines docker and other devops shit

But i still feel like im missing on A Lot of stuff regarding aws. I havent worked with Fargate for example and dont know how it works or when to use it, but i heard other devs use it

Can someone list me a number of things i as a dev should know more regarding aws?

Question:

For a real time chat (web) app. Whats the best technology to use for this? I dont want the chats to have delays or glitches (in case it gets sent but not delivered etc).

Backend stack is java spring boot

How about kafka? Or rabbitmq? Or socket.io? Should i use redis? Should i use AWS SQS? Talking about cloud what AWS tools should i be using to handle this the best way?

Note: it must be scalable. Meaning if i wanted to extend this software more by building a mobile app (aside from web), i should be able to use the same backend easily

Note 2: it is not ONLY a chat app. Chatting is just 1 out of many functionalities. So chatting is not the main component of the project just a side thing

Keep in mind the backend is a microservice architecture etc. Database postgresql.

So I am pretty fair dev at Java and have been doing freelancing for sometime apart from normal full time job.
Got a client , a well funded one, who raised a decent chunk of money recently.
Got me do a couple of different areas right from refactoring and bumping their performance to all the way setting up AWS Services like RDS,Lambdas,Dynamo,SQS.
It was going good , money was coming in for the initial part.
Thinking that money is not the concern here , I accepted work at runtime and gave quotations about the additional work.

However now that all is done and deployed , the client simply refuses to pay me the money and has ghosted me horribly than my ex ever did.

I have access to their GitHub,AWS(I setup myself).

Need suggestions of whats the best way I can fuck them up if they decide to not pay even after a few more professional polite attempts I do .

sidenote : They had a pretty dumb db design and blindly had resorted to services in AWS and the pricing is still a major point of concern for them.