Serverless build stuck at "Creating cache directory" (infra issue?) | Runpod | Page 1

lavish owl Jan 13, 2026, 4:46 AM

#

Hi, I’m seeing repeated serverless build failures and wanted to check if this is an infra issue.

Endpoint ID: aheqed88722qgo
Issue: build worker fails at “Creating cache directory” / pod not created
Tried: cache disabled, rebuild, same result
Context: works locally, fails only on RunPod serverless

Is anyone else seeing this today, or is there a known issue/workaround?

stuck crownBOT Jan 13, 2026, 4:46 AM

#

loud skiff Jan 13, 2026, 3:51 PM

#

Hi there, are you still facing this issue? could you please post the build id here. Thank you!

lavish owl Jan 14, 2026, 2:23 AM

#

My last build id is "da5ca801-49c2-488c-83c0-95cfa3a5c03a".
I'm still facing this issue... Thank you

quaint wadi Feb 12, 2026, 10:32 AM

#

Exact issue as OP 😭

#

Im gonna have to do a work around for now but would be nice to know why thtis may be happening

#

Cant give much more info as the build logs are providing very limited information

shut walrus Feb 12, 2026, 11:54 AM

#

Most of the time it's Dockerfile issue like to many RUN layers

hybrid charm Feb 12, 2026, 1:20 PM

#

quaint wadi Cant give much more info as the build logs are providing very limited informatio...

Last time I checked the vLLM 2.11.3 release failed due to a mismatch between the tokenizer and transformer architecture.

unique scarab Feb 12, 2026, 1:58 PM

#

Same issue here. Builds fine locally, and also built fine with the same number of RUN layers on Runpod before...

#Serverless build stuck at "Creating cache directory" (infra issue?)