Lots of resources and large configmaps cause tilt to hang

[glightfoot]

Expected Behavior

Tilt should not hang and become unresponsive

Current Behavior

Tilt hangs and becomes unresponsive to any actions. The tiltfile loads, but show status stuck in updating.

Debugging shows the main goroutine being stuck on:

goroutine 1 [sync.RWMutex.Lock]:
sync.runtime_SemacquireRWMutex(0x1401a090d78?, 0x4?, 0x14000c52501?)
        /usr/local/go/src/runtime/sema.go:105 +0x28
sync.(*RWMutex).Lock(0x1401a091468?)
        /usr/local/go/src/sync/rwmutex.go:155 +0xfc
github.com/tilt-dev/tilt/internal/store.(*Store).Loop(0x1400072be60, {0x104ac6380, 0x14000c4fb90})
        /root/project/internal/store/store.go:161 +0x1d8
github.com/tilt-dev/tilt/internal/engine.Upper.Init({0x10466fd40?}, {0x104ac6380, 0x14000c4fb90}, {{0x14000b55b30, 0x42}, {0x106d684e8, 0x0, 0x0}, {{0x103ef7d38, 0x6}, ...}, ...})
        /root/project/internal/engine/upper.go:102 +0x58
github.com/tilt-dev/tilt/internal/engine.Upper.Start({0x2?}, {0x104ac6380, 0x14000c4fb90}, {0x106d684e8, 0x0, 0x0}, {{0x103ef7d38, 0x6}, {0x0, 0x0}, ...}, ...)
        /root/project/internal/engine/upper.go:88 +0x178

and a large percentage of time spent on apicmp.DeepEqual while holding a RLock on the store mtx.

Steps to Reproduce

1. Spin up large tilt environment with 500 resources, using lots of file watches
2. Create large configmaps and load them into tilt
3. trigger a watched file change to retrigger the bug

I have also included a tar.gz file with a setup that seems to reliably reproduce the issue. I used a kind cluster for that setup, as shown in the tilt doctor output below.

Context

tilt doctor Output

$ tilt doctor
Tilt: v0.35.0, built 2025-06-13
System: darwin-arm64
---
Docker
- Host: unix:///Users/greg.lightfoot/.docker/run/docker.sock
- Server Version: 28.3.2
- API Version: 1.51
- Builder: 2
- Compose Version: v2.38.2-desktop.1
---
Kubernetes
- Env: kind
- Context: kind-tilt-repro
- Cluster Name: kind-tilt-repro
- Namespace: default
- Container Runtime: containerd
- Version: v1.27.3
- Cluster Local Registry: none
---
Thanks for seeing the Tilt Doctor!
Please send the info above when filing bug reports. 💗

The info below helps us understand how you're using Tilt so we can improve,
but is not required to ask for help.
---
Analytics Settings
--> (These results reflect your personal opt in/out status and may be overridden by an `analytics_settings` call in your Tiltfile)
- User Mode: opt-out
- Machine: 323e9d1ad6016ab4822de03a28c50b84
- Repo: 3y4NopketDVRWDViKN9sPA==
...

About Your Use Case

We use tilt extensively for developer environments, and some of those are quite large. We have seen an increasing number of problems with tilt becoming stuck, requiring the developers to have to recreate their environment.

[glightfoot]

here is the repro file. I had cursor add random stuff to the tiltfile until it reproduced the bug, but it seems to be linked to the large configmaps as those are what finally triggered it.

tilt-lock-repro-small.tar.gz

NOTE: I was running tilt on port 10351 since I had our prod tilt running and stuck on the main port, so the collect-diagnostics script references the 10351 port