Studio Unloading Assets Due to HTTP 429 too many requests On One Specific Account During Playtest

I logged onto studio this morning and was greeted with what I can only assume to be a bug. I have attempted uninstalling all of my plugins, disabling API services, ETC, and I have not solved the problem on my own.

When I load a playtest in studio, it will start revoking assets (textures, sounds, animations, meshes, etc.) rapidly under HTTP 429, claiming it cannot load them due to excessive requests, which I don’t understand because as demonstrated in the video, it loads them fine beforehand before unloading them. Worth nothing, this ONLY occurs in Roblox Studio, not in regular on-platform playtesting after publishing (and I can publish just fine.)

There are no scripts in game that make ludicrous amounts of requests, and HTTP requests are disabled.

I at first assumed this to be the fault of me or a plugin accidentally pushing the server or whatever too hard, but I waited a good 15 minutes out of studio, loaded other studio files, etc. and I cannot reasonably assume this is just standard API rate limiting for one experience. It occurs across all of my games no matter what.

When testing on an alt on the same experience in team create, I experienced none of the issues. It is solely relegated to my account, @nekstoer. This bug also does not occur if I start a Team Test, so I’m assuming it has to do with the fact that it’s my account that hosts the server in question. Also noteworthy: this does not occur in unpublished instances (IE: If I were to download a copy of the place without saving it to Roblox as an experience, I experience no issues loading everything)

During play, I notice that the amount and what assets aren’t loaded are all completely random every time.

If by some chance this is not actually a bug and I’ve mistakenly posted it here under that impression, I apologize, but no other developers who I’ve consulted for help know what’s happening either, so I can really only assume this is a bug.

Expected behavior

Studio would load all assets without issue.

8 Likes

For any staff who may read, I did not include log files in the original post by accident. I have them ready but I am unsure where to add them privately in post.

I am experiencing this exact same issue but on both local and published places…

2 Likes

I think it might have been a beta feature? I tried disabling them all again and the issue is mostly fixed, but I’m still having minor ratelimiting issues.

Same issue here. I wish I knew what to do to make this stop happening. It’s making testing very difficult.

3 Likes

I’ve been suddenly having this happen over the last few weeks.

sometimes playtesting / stopping studio will load everything or almost everything, and other times almost everything will unload with the “Too Many Requests” error.

Seems to maybe happen way less often in team studio places if someone else is in studio with you?
Meanwhile places I’m working on by myself fail to load stuff every other playtest.

1 Like

How long is the interval between times where studio loads everything and doesn’t? Since I made this post I’ve been able to load stuff pretty consistently (i think, i haven’t been in any asset heavy places), but I just noticed it tanking after playtesting a game with a lot of assets.


also, also, i’ve been getting this warning? May be related. No clue what it means.

It’s super inconsistent.

When I made that post it was happening anywhere from every other to after every two play tests.
That lasted a couple days, though again editing a team place with someone else in it seemed to reduce the frequency by around more than half.

Today it hasn’t happened at all.
I suppose it could’ve just been poor connection over the weekend, but its happened so frequently over the last month it could also be a bug, or both.

Same. It stopped happening for days and then it started again today.
I don’t know if it’s just a spurious correlation, but it happened after using my VPN for a bit, while I left studio running. I Cut off the VPN, but after reconnection, the issue started again. It may just be a coincidence, though.