Hacker News new | past | comments | ask | show | jobs | submit login

Yes, I was considering Hetzner or Scaleway, I have a few months to decide before my current prepaid plan expires.

Thank you for the heads-up !




I've had 0 issues with Hetzner's servers, they're great at that. I hoped this would translate to cloud services as well, but it turns out (not surprisingly in hindsight) that offering a fully managed service is something completely different than cheap and easily administered server.


That's a good point. I've been surprised and frustrated by the S3 issues we've encountered since the hundreds of servers we are hosting on Hetzner Cloud have rarely had any issues. Only now i realize that this is the first service they are offering apart from networking that includes LB.


They launched recently their S3 Storage, so I think some initial issues are inevitable.


True, but then they shouldn't claim "High availability" as an empty promise without even a non-binding expected number, like 99.9%.

Also, it's not good that they do not have any status tracker that shows summary stats on uptime over the past, say, year so one can get a quick idea of how reliable the service was historically.

The status monitor only has entries for incidences from the past 7 days. It could have been down for 10 days a month ago, you wouldn't know this unless someone wrote about it.

_You_ know the service is new, but this isn't mentioned on the landing page either. They don't say something like "only 4 months old, public beta" - in that case I wouldn't feel negatively surprised.


I just had a look at their page. Their S3 is different from Amazon. In AWS, it is global - they synchronize the objects between locations which menas it's eventually consistent but it has amazing SLA. Whereas Hetzner clearly states: "Currently, Object Storage is available in the following locations: Falkenstein, DE (FSN1), Helsinki, FI (HEL1), and Nuremberg, DE (NBG1)." So if I wanted HA, I'd use two locations or more.


As someone who has been managing his own servers (and also hosting services for his clients) for many years, I can understand the issues, problems and difficulties that can arise, so I have learned to be much more patient and understanding.

I learned (the hard way) to have a plan b for everything. I'll study an emergency solution in case of a prolonged downtime.




Consider applying for YC's Fall 2025 batch! Applications are open till Aug 4

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: