VRChat has now been down for several hours on Christmas Eve, from about 8PM EST to now about 12 AM. This is absolutely unacceptable. I don’t know the rules of this forum, and I don’t care. All I can tell you is that as a person that has been involved in IT for 30+ years this is absolutely ridiculous. Even minutes of downtime at most major companies is unacceptable. This is hours we’re talking about. Many of my friends have not been able to get together for Christmas Eve because of the incompetence of this company. I have been a VRChat supporter since VRChat+ was a thing. I’ve played since late 2019, and I’ve been a supporter. All I have to say is shame on you all VRChat team. I have praised you, I’ve written good reviews for you on teams, I’ve vouched for the dev team even when all my friends were upset because mods were taken away. But not today. Today I can say for certain you guys are absolutely failing. This is a social platform with real people and real social connections you failed to help connect for the holidays, and I hope that is on all of your minds for Christmas. That is all I have to say. Be better, do better.
Sorry, an upstream provider’s services were down, which affected VRChat and many other applications. We had a full team on staff, online & monitoring, and waiting for their services to come back online and inform our users. We understand the impact that even a moment of VRChat downtime has on users, and we do everything we can to minimize it – but sometimes, it is unavoidable.
We’ll pass on your feedback to our provider!
The upstream providers are kind of the problem and you guys already know this. The tech debt is really starting to show through and there is going to be a lot more of this if you don’t recognize the weight of what it is we’re actually facing as internet users right now.
This isn’t remotely vrc specific, but vrc IS in the best position to be ahead of the coming wave.
You had to know this was coming given the last few weeks of issues the platform has been having Noxy… It’s going to happen again in about six days, in case you’re wondering (new years).
An upstream provider for vrchat is probably a technical component like a cloud provider, it’s not related to “technical debt”, it’s a strategic choice to not have to do infrastructure in house that costs millions.
I believe that if vrchat is down for a few hours we will survive. But i imagine, but if you rely only on vrchat for all your social connection, that’s always going to be a risk, and you will have that on any platform that you are not owning or managing yourself.
I would ask that you stick to your own profession rather than making claims you seem to lack comprehension in.
Relying on SaaS because you’re too lazy to do it yourself is the literal, objective definite example of technical debt.
Relying on cloud providers was fine TEN YEARS AGO when VRC was first spinning up. They’ve had the time and the incentive to move on and build their own solutions by now.
I have been a software engineer for years, I know how the industry works.
Sorry to disagree, but the proportion of people that do their own hosting is very minimal nowadays.
Here I believe vrchat uses AWS but I may be mistaken. It’s probably (can’t say for sure) used for hosting servers that are making vrchat run in the background.
I would be very surprised if they were hosting those servers on premise, in their garage 
.
LOL
VRChat has been down in the South American region for the entirety of December at this point and VRChat doesn’t seem to give a single fuck about it despite a lot of users reporting.
Ihad to download Cloudflare Wrap as an alternative solution to being able to connect and the solution came from a random user, not from VRChat team, they don’t seem to care at all.
You must be entierly unaware of the homelab revolution trending across multiple social media platforms for the past few years.
Yeah hopefully the crowdstrike, aws, and let me see, oh around six other major saas provider outages this year alone which have wiped out well over half the entire internet, irreperably I might add, will change that shortsighted and idiotic mindset. Who knows though, it’s anyone’s guess what the future holds.
I’m reminded of that statement your parents told you, about if everyone is doing something: Would you jump off the bridge too? I self host my entire stack from ground up using fully open source software. As soon as I can get a webxr endpoint that let’s people use their own avatars setup and deployed on my server without crashers, I’ll have replaced vrc too.
You also explained the sole reason they lack profitability, even with significant conversion on subs: Overpriced and useless cloud servers.
This is a hard take, but realistically the VRChat team can’t do everything.
Yes, they could self-host their own datacenter. That would mean owning and maintaining globallaly located servers, hiring 24/7 teams, managing redundancy, DDoS mitigation, regional compliance, power, hardware refresh cycles, and disaster recovery. Those costs do not disappear, normally this is for AWS or the company hosting the hardware, and because they do it on-mass, it’s cheaper..
The outcome would be either a significantly more expensive VRC+, aggressive monetization (ads), or a slower development course for VRChat. There is no alternative that the VRChat team can use that is also feasable monitary wise.
Relying on cloud and upstream providers is not technical debt. It’s a trade-off used by the overwhelming majority of just about any platform, from facebook, to twitter (x). Even companies that do self-host still depend on upstream providers, CDNs (Fastly, Cloudflare, etc), DNS (Namecheap, Cloudflare Registrar). There is generally nothing like full independence at VRChat’s scale.
Could services be better isolated or regionally offloaded to reduce the effect of downtime? Absolutely, that’s something you could discuss about. But framing cloud usage itself as laziness or incompetence ignores how platforms like VRChat are actually operated in this day and age.
Homelabbing is great. And should absolutely be encouraged for learning, experimenting, and running personal or small projects. But at VRChat’s scale this is an an entirely different problem space. If you want to run something in production for millions of users, it needs to be properly developed, secured, monitored, and supported accordingly.
I saw some users have that kind of issues. Usually it’s connectivity issues to their service, not their service itself. But I suppose.good that you found a solution that works for you then.
Ready yourself for the future from the past: web . janusxr . org 