Has AWS support gone downhill in the last two years? I've worked with them in the past - as both an individual and a couple startups - I always reached a human. Issues weren't always resolved as quickly as I'd like but response times were short.
Working in small and medium businesses I've observed the same thing, and I've been quite satisfied with it. So I don't think it's really gone downhill, so GP's comment doesn't really resonate for me, but that isn't to negate their experience. Otoh I keep hearing horror stories about GCP and now I'm reluctant to try it.
Shitting on GCP is just popular on HN and always gets upvoted. AWS and Azure have royally fucked thousands of customers if you care to search for those writeups. My wild ass guess, considering posts like these have zero background details, is that they were careless with service account keys and their account got suspended for mining crypto or something. They also probably weren’t actually paying for support of any kind and that’s why no one is responding to them.
I have the same question. I always got human response after 24-48hours or after one round of messages (with an automated human or machine, not sure). But so far, across 3 accounts and a dozens of correspondence, I always got a human.
In the 2010s I always got an AWS support team to help.
Now I get handed off to an external partner of AWS certified contractors.
They are often terrible. They have no backend systems access and just run through the AWS equivalent of "reboot it", "defrag your disk". Basically trying to find an issue in my pipeline. Which they never do because it's the same TF scripts used for years.
Only once we waste time going through the motions do I get passed up to someone who can actually correct the backend issue in the AWS stack itself.
Tbf though I rarely ever have to contact AWS support at this point. The few times I have in the last 2-3 was due to issues after they rolled out an update or with a newer service we wanted to use.
Never have issues with stable services like S3, ECS, EKS, or RDS.
My last experience was with someone giving me answers that where clearly in part LLM generated: It contained payloads/configs that did not even match the actual API
> They have no backend systems access and just run through the AWS equivalent of "reboot it", "defrag your disk"
To be fair I would bet money that the overwhelmingly vast majority of support tickets are exactly those kind of issues, and ones that refer to actual bugs on their end are, comparatively, extremely rare, and should have to be escalated through normal procedures to weed out common problems.
politelemon|29 days ago
vrosas|29 days ago
ranger_danger|29 days ago
csomar|29 days ago
EdwardDiego|26 days ago
But that's only based on a bunch of anecdata.
benSaiyen|29 days ago
In the 2010s I always got an AWS support team to help.
Now I get handed off to an external partner of AWS certified contractors.
They are often terrible. They have no backend systems access and just run through the AWS equivalent of "reboot it", "defrag your disk". Basically trying to find an issue in my pipeline. Which they never do because it's the same TF scripts used for years.
Only once we waste time going through the motions do I get passed up to someone who can actually correct the backend issue in the AWS stack itself.
Tbf though I rarely ever have to contact AWS support at this point. The few times I have in the last 2-3 was due to issues after they rolled out an update or with a newer service we wanted to use.
Never have issues with stable services like S3, ECS, EKS, or RDS.
Garlef|29 days ago
ranger_danger|29 days ago
To be fair I would bet money that the overwhelmingly vast majority of support tickets are exactly those kind of issues, and ones that refer to actual bugs on their end are, comparatively, extremely rare, and should have to be escalated through normal procedures to weed out common problems.