>AWS: outages>Azure: hold my beer
>>107043687Why would anyone use these cloud services? They're much more expensive, less reliable, and harder to use (because you have to learn a bunch of proprietary software).
>>107043687These guys probably weren't doing anything important, right?
>>107043701simple: before, people would shout at me to fix itnow, I get to just shrug and say "not my fault boss, MS shat the bed (again)"
>>107043687>amazon cloud fires a bunch of indians>widespread outages>microsoft cloud fires a bunch of indians>widespread outagesPerfect for terrible look saars
Time to get comfyDon't fix it too fast MS, take your time. Heck, take a long weekend even.
lol
>>107043701Because IT incompetence is usually worse than cloud incompetence. Usually cloud shit will work as expected and can be built up on-demand.
>>107043687AAAHHHHHHHHHH I CANT WORK AND MR SHEKELBERG IS LOSING MONEY AHHHHHHHHHHHHHHHHHH FIX IT MICROSOFT
>>107043701It's all just blame shifting.>IT shifts blame to cloud because the management wanted it>management shifts blame to CTO because he said cloud migration NOW>CTO shifts blame to consultants who said go to the cloud everyone does it>cloud provider says something something DNS and everyone is happy
>>107043831This. It simply isn't my problem anymore.
>>107043858Except that's not true because you're the first to get roasted by your company over and outage out of your control and they don't give a fuck that you outsourced it.
>>107043831It's easy decision for business to prefer hosting with MS and their Indian army who may have seen outrage once in a whole rather than trusting that creepy anon who gives everyone the ick and put the company in the shit because he put all the production servers on fucking Gentoo and noone could figure out how that shit worked after anon got arrested at work for downloading terabytes of illegal my little pony cartoons
>>107043867As they say, no one ever got fired for buying IBM.
>>107043868How do I unsubscribe from your self-loathing diary?
>>1070438701) that's definitely not a lie2) outages are miserable so your day is ruined
>>107043687winget was failing to pull updates yesterday...I figured that was a preamble for another MS breakage Related, AWS is having issues again today too.
>tech for full jeet and AI totally not for saving money>suddenly everything starts breaking
>>107043888Seems it's Front Door services being affectedThat's a global service, and all those people who planned for regional failures by having a warm standby in another region are also fucked. Woops.
https://azure.status.microsoft/en-us/statusIT'S OVER
SAAAAAAARS SEND HELP AAAAAAAAAAAAAAAAAAAA
Oof! Ouch! Owie!
I would love it so much if there were daily cloud platform outages for the rest of time. Please God keep them coming.
what's even the point of azure? More exoensive than aws in every way. Als the UI is just as bad if not worse.
>>107043868You joke about this, but I worked with a nigga who actually used Gentoo on company equipment and I would install EL7 over it because I couldn't be bothered to fix the autismally minimal kernel config that didn't have the drivers I needed for the job. Nigger faggot died of heart disease a few months later. Good riddance.
>>107043994being locked in to one particular ecosystem is the biggest reason to choose one particular platform over the others.
>>107043687if azure, aws, and cloudflare go down the internet is gonewhat happened? why are we like this now?
lolle the company I work for uses azure and the customers are going batshit. I just disconnected my phone ecks dee.
>>107044005Because of economies of scale. They can offer the cheapest servers at the lowest distance.
>>107043994I chose it at work because Azure SQL was less autistic than dealing with SQL Server PaaS. My only other DB choice was Oracle which requires nonfree drivers built to dynamically link a specific libc or their faggy jar jdbc driver. Basically every language I looked at has a working TDS 7.x driver that's free and open source. Fuck Larry Ellison.
>>107044005Just need to pump billions more into LLM's bro it'll solve everything promise
Stinkjeets are to blame for this
>>107044026Fair enough. I dindu no SQL stuff so I guess there are some use cases for azure, no pun intended. In terms of gpu compute pricing, they simply do not want to be competitive.
Outages are the best time everSelf-hosting fags could never understand
>>107043964>>107043970
>>107044101Shop fronts don't want to use Amazon, their competitorTemu, Walmart, etc.
>>107043701because you need to sue someone if shit goes down. Economy is a non stop loop of risk shifting because you don't want to be responsible for someone's lost revenue.
>>107043687don't worry, it will be much worse in the future
>tfw US Gov Cloud still worksAAAAAAAAAAAAAAAAAAAAAAAAAAIT'S NOT FAIR
>>107044005Jews
>>107044045https://www.youtube.com/watch?v=VTmJJ0lUyQc
>>107043701Because muh offsourcing muh cheaper muh redundancyBut as proven by AWS and now Azure going down: None of that shit matters when you want 99.999% uptime or get reamed in business profits.
>>107043687Imagine using *neNote when Notepad++ is in your software center, now those cucks can't work because all of their shit is in buggy onedrive limbo maintained by thirdies. Whevnever theres an outage i have to wonder if they're just getting hacked. anytime they get hacked or exfiltrated surely they call it an outage as well?
>>107044633As if any small company can provide 99% uptime. They are just as jeeted as anyone else and will have shit qos as well. The only difference is it will be more distributed when it happens
>>107044005Poop skinned people
>>107043687On premises wins again, why did we migrate again, boss?
>>107044727see>>107044354
Tommy G kino
>Pinging microsoft.com [13.107.213.69] with 32 bytes of data:>Request timed out.>Request timed out.>Request timed out.>Request timed out.
>>107043996If you're gonna use Gentoo in production at least make it a binary only install and/or build the binaries elsewhere.that said, I'm not sure you could completely remove the compiler so you could get some exploits there since Gentoo will always have a compiler ready to be called
>>107043753>t. Cloud Pajeet
>>107044005Boomers love for the Poojets clouded their minds.
>>107043906Come on anon you can't seriously be proud of that road design
>>107043687AI will prevent this.
>>107043710>Fire 9000 white employees>Hire 10000 jeets>Outages become commonplace on platforms that used to boast about having 99.9% uptimeComplex systems will not survive the poos festival
>>107044677>They are just as jeeted as anyone else and will have shit qos as well.t. JeetI don't disagree that 5-9 uptime is nearly impossible, but you're dumb if you think not having your shit on-prem isn't head and shoulders better for situations like this.>Oh shit, Dallas has a cut pipe?>Redirect everyone to New York/Seattle/LA/Chicago!
Has WWIII just begun?
>>107044677It’s pretty trivial to do. have you ever actually built servers + server racks? Or are you just a coonsumer?
>>107043701Retards falling for sales pitches. I swear I could've built a house out of free fucking Yeti mugs by now if I spent the time to listen to every cloud salesperson who offered me one.
>>107044909Fuk u saar my vilag took 500 Brahmin and twenty days to get server on.
>>107044909I work at azure
>>107044936Saar the system is down please do the needful and kindly fix the servers
>>107044633To be fair, all Amazon and MS documentation explains how you're supposed to do multi-region setups for things you want serious uptime on. Having all three of your kubernetes nodes sitting in us-east-1 is not redundancy just because you used different availability zones. Hell, I don't like doing anything in us-east-1, it's so damn big it's doomed to have problems.
>>107043687Why do cloud outages seem to happen together like this? It seems like one will have a big outage and the other will do the same thing a few days later.
>>107044800this one's much better
>wanted to download VS CodeFucking freetards can keep up Codium instances longer than M$.
>>107044976One tectonic plate shift bumps into the other. Rippling instabilities after any outage make a second outage more likely, at any scale. A shitload of people massively moving stuff from AWS in a panic to Azure could have thrown extra load on there, some people probably had systems that failover between cloud providers.
>>107044969Not my team, don't care
>>107044972this is affecting Microsoft's CDN provider (azure front door) that's why so many MS sites are down right now. MS official advice right now is to change your DNS to point directly to your origins rather than via their global CDN/reverse-proxy service. So even if you diligently set up your infa to have failover occurring to MS guidance and paid the extra for having that failover, your still being completely hosed right now
>>107044982>Resurrection Cemetery
>>107045018
>>107044757is it just me?https://www.youtube.com/watch?v=MN3x-kAbgFU
>>107045027>King was writing about a real placeDose this mean a clown haunted Maine in the 50s?
>>107045033Microsoft.com won't even load for me
This is a real non-ai generated road designed and built in a certain country (you'll never guess which one)
My cluster exists in multiple datacenters.
>>107044936So, no. Understood
>>107044824azure never had 99.99% uptimeits the worst cloud platform imaginableonly retarded businesses get duped to use this garbage because of "microsoft incentives"
>>107043720this. I use proprietary software strictly because of this.Foamers can't comprehend the concept of not wanting to overwork yourself for no reason. If it breaks it breaks, who cares the world is better without needless terminally phone bound addicted faggots anyway.
>>107044677So wrong.
>Starting at approximately 16:00 UTC, customers and Microsoft services leveraging AFD may have experienced latencies, timeouts, and errors.>AFDextreme left has mixed Azure Front Door and Alternative Für Deutschland and they are now denazifying the internet?
>>107045099>the paperclip bow and arrow tribal battles during the Great Azure Outage of 2019My peoples in Operations held the line against the BizReqMgmt tribe's assault for lo two days
>>107043687on prem chads would be winning except they'll do the math and realize they still have more down time and shut up before bringing more attention to this
>>107045201kek
If I was Huawei I'd be partying right now
>>107045754>more down timeseriously doubt thisalso at this point AWS/Azure have had insane downtimes you can do literally nothing aboutservers can literally go decades without a single update or change and NEVER crashbut whoops, some jeet nigger at amazon dropped a . in the DNS
>>107044909how do you deal with some cracka ddosing your shit?or if your server room catches on fire? 99%?
>>107045935if you have DDOS problems you set up a POW mac mini bootcamp proxy.servers don't catch on fire, just like your desktop and laptop doesn't catch on firealso imagine betting that AWS isn't so incompetent that some Jeet doesn't whoopsie doopsie delete a trillion files one day, tick tock
>>107043701>buy space in a datacenter>pay for cabling and installation of hardware>pay rent and security>management patching of OS>administrate the machine>go through entire process again when you run out of capacityvs>hit deploy
>>107045979>pay $10,000 for $1000 in services and still goes down because they hire Jeets>implying that 99% of what's hosted on AWS requires more than a Mac Mini.
>>107044727Our bosses attended seminars sponsored by the big cloud providers talking about all the upsides while conveniently leaving out the real downsides.
>>107043701You don't need an in-house IT department if someone else's computer will do it for you.Tech workers get fired, the industry becomes a monopoly, then suddenly nothing works and we all wonder why.
>>107046033unironic given AWS requires someone to manage it, it's one of the most convoluted and annoying hosting options out there, if you fire your IT department you're even more fucked with AWS because you need a nerd to even log in let alone manage anything
>>107045999Fucking this. Big data, cloud solutions, AI, it's all the same shit purple dragon chasing to justify change for the sake of change.>>107046033As long as office workers still use laptops and in-house software, a company still needs an in-house IT department.
>>107046132>As long as office workers still use laptops and in-house software, a company still needs an in-house IT department.Logic would dictate that. But there are external providers that offer this service.It works about as well as you'd think.
Damn, they really really dropped the ball this tirme
97% SLA? What the hell do you need 96% SLA for? Come on, just because we advertised 99.99% that doesn't mean we can provide the ridiculous 95% number you're asking...
>>107045985>not my problem because everyone blames the platform now
>>10704627999.99% uptime? Who would need 80% uptime? I don't have 50% uptime! Help a madman is trying to kill me over my last 10% uptime! You'll be hearing from my lawyer about that broken 99.99999% uptime contract!
>>107046194>azure healthcare apiThese niggers are going to get someone killed.
>>107046303No need to get that low, 94% means 21 days offline per year.How far away from that are cloud providers anyway, lol. It's incredible how shit they are.
>>107046194They should have listened to him warned you about microsoftall that shit about ransomwareazure is the biggest ransomware
>>107043687Just disable Copilot, bro.
>>107043701It’s a marketing trap. But also a job guarantee since everyone’s in the same trap. You got a couple cloud engineers or "DevOps" that lobby for AWS or any other hyperscaler, NaiveDate managers that write down some decision report littered with logical fallacies, and a few years in the sink cost is so high you can’t get off of it, and instead of doing productivity work you’re sitting in myriads of FinOps meetings, where even fewer understand what’s going on.Engineering mangers are promised cost savings on the HR level. Corporate finance managers are promised OpEx for CapEx trade-off, the books look better immediately. Cloud engineers are embarking on their AWS journey of certification being promised an uptick to their salaries. It’s a win/win for everyone, in isolation, a local optimum for everyone, but the organization now has to pay way more than it—hypothetically—would have been paying for bare metal ops. And hypothetical arguments are futile.And it lends itself well to overengineering and the microservices cargo cult. Your company ends up with a system distributed around the globe across multiple AZs per region of business operations, striving to shave off those 100ms latency off your clients’ RTT. But it’s outgrown your comprehension, and it’s slow anyway, and you can’t scale up because it’s expensive. And instead of having one problem, you now have 99 and your bill is one.
Msft doesn't care about you. The big dogs are openai, Walmart, govt, black rock, etc. Nobody cares that some no name bookstore in Ohio lost connectivity. Boo hoo, that Walmart contract is still $100b
