Previous the space what took 1min to start. Now taking 1hr in build and 1hrs in container. Idk why… I don’t get any log. I restarted, factory reset it nothing works. It’s effecting all of my spaces new and old.
Hi, it looked fixed yesterday afternoon, but now it looks like it is not working again, my space is stuck in bulding for no reason (I even tried to push again the same files of the version that is currently running), do we have more information about it ?
I am also having the same issue, rebuilt my old space, and it is stuck on “rebuilding”. I created another space and added files, it is still stuck on “rebuilding”.
I also notice that when the build eventually completes, the new version starts serving and works, but the HF infrastructure doesn’t notice and still shows it as Restarting.
is there any chance someone from HF is looking at it ? do we have some other place to inform them? I feel like I am wasting time since last 4 days on it.. just watching “building”. Think about we are unresponsive to our customers or stakeholders?
Hmm… Major official HF updates usually take place during weekday daytime hours in the US. This particular issue was significant and the timing was unfortunate… Even on holidays, staff sometimes respond semi-officially out of goodwill. (For example, hysts posted a message earlier.)
The shortcomings of HF’s announcement system (in a conceptual sense, not referring to specific pages or functions) have existed for quite some time, but we haven’t seen any decisive improvements.
I’m not exactly a veteran user of HF, so I don’t know the background or “why things are the way they are”…
BTW, here’s an example of a feature added to the announcement system in recent years: Changelog - Hugging Face
Edit:
As for features that have been around since I first started using HF, the HF Discord and Posts are sometimes used for announcements.
I have the same problem. My Space is stuck on the blue startup screen, and after a while it gives me this error:
runtime error
Launch timed out, workload was not healthy after 30 min
I’ve tried everything, but nothing works… And to think it was working just yesterday…
Launch timed out, workload was not healthy after 30 min
Hmm… Given the timing, the infrastructure might be to blame, but this error often occurs for one of the following reasons:
Dependencies have been broken—perhaps due to an update to a library online—causing the installation to enter an infinite loop that cannot be resolved even after 30 minutes
The build simply includes a library that takes longer than 30 minutes to complete
The port used to launch the GUI’s backend process is incorrect, preventing the backend and frontend from communicating even after 30 minutes