1.6.0 fatal error on launch

[PhilETaylor]

Describe the bug

Saw your tweets, upgraded to 1.6.0, started Orbstack - got "Stopped unexpectedly: fatal error"

To Reproduce

upgraded to 1.6.0, started Orbstack

Expected behavior

no error :)

Diagnostic report (REQUIRED)

🌲 vmgr | time="05-13 11:33:41" level=info msg="using proxy: none"
🌲 vmgr | time="05-13 11:33:41" level=info msg="forwarding SSH agent" sock="/Users/phil/Library/Group Containers/2BUA8C4S2C.com.1password/t/agent.sock"
🌲 vmgr | time="05-13 11:33:41" level=info msg="starting VM"
2024-05-13 11:33:41.159 OrbStack Helper[94926:2603733] [brnet] guest read error: 54
2024-05-13 11:33:41.159 OrbStack Helper[94926:2603726] [brnet] write error: 39
2024-05-13 11:33:41.159 OrbStack Helper[94926:2603726] [brnet] write error: 39
2024-05-13 11:33:41.159 OrbStack Helper[94926:2603726] [brnet] write error: 39
2024-05-13 11:33:41.159 OrbStack Helper[94926:2603726] [brnet] write error: 39
panic: Internal error while starting microVM: Vm(SetUserMemoryRegion(MemoryMap)) [recovered]
	panic: Internal error while starting microVM: Vm(SetUserMemoryRegion(MemoryMap)) [recovered]
	panic: Internal error while starting microVM: Vm(SetUserMemoryRegion(MemoryMap))

goroutine 1 [running]:
main.runVmManager.func3()
	github.com/orbstack/macvirt/vmgr/main.go:462 +0x60
panic({0x10443bac0?, 0x140002721a0?})
	runtime/panic.go:920 +0x26c
github.com/orbstack/macvirt/vmgr/util/errorx.RecoverCLI()
	github.com/orbstack/macvirt/vmgr/util/errorx/errorx.go:29 +0x78
panic({0x10443bac0?, 0x140002721a0?})
	runtime/panic.go:920 +0x26c
main.check(...)
	github.com/orbstack/macvirt/vmgr/main.go:93
main.runVmManager()
	github.com/orbstack/macvirt/vmgr/main.go:739 +0x2c20
main.main()
	github.com/orbstack/macvirt/vmgr/main.go:977 +0x78


Check logs for more details.

Screenshots and additional context (optional)

[kdrag0n]

Please share a diagnostic report from Help > Upload Diagnostics or orb report.

[PhilETaylor]

OrbStack info:
  Version: 1.6.0-rc1
  Commit: 932a523a3e455516178ac5c35f345965de2138de (v1.6.0-rc1)

System info:
  macOS: 14.5 (23F79)
  CPU: arm64, 20 cores
  CPU model: Apple M1 Ultra
  Model: Mac13,2
  Memory: 128 GiB

Full report: https://orbstack.dev/_admin/diag/orbstack-diagreport_2024-05-13T10-36-37.510498Z.zip
<!-- (To review the report, check ~/.orbstack/diag) -->

[PhilETaylor]

drats - now when I try to install the stable version I then get a different error...

So I deleted (renamed) ~/.orbstack, downloaded stable version, launched it, worked, upgraded to canary and it launched with no issue. So the problem with ugprading is somewhere in my .orbstack folder :)

Hope that helps.

[kdrag0n]

It's the memory size. Will fix but in the meantime it should work if you set it to ~60GB or less.

[PhilETaylor]

No worries.

[kdrag0n]

Fixed for the next canary.

[PhilETaylor]

Awesome - thanks.

[kdrag0n]

Released in v1.6.0 Canary 2.

[PhilETaylor]

Upgraded... no further issues - thanks again.

[kdrag0n]

Released in v1.6.0.

New: Truly fast container filesystems on macOS: 2–10x faster, within 75-95% of native

[PhilETaylor]

Thanks - had for the last few days a few issues with your internal networking but not mentioned it as I cannot replicate each time, I hope its not a bug but I think it might be. *.orb.local domains seem to be routing to different services or ports - hard to replicate to give you any detail at this stage, but orb stop/start and then docker-compose up again seems to fix it. This only started in the canary for 1.6. Not had it yet on 1.6.0 as released so will keep an eye on it (unfortunately I switch projects a lot during the day haha)

[kdrag0n]

Known issue, will fix in the next version.

[PhilETaylor]

ah man you are a star! I did not want to mention the issue until I had rock solid proof of concept to replicate it for you.

Awesome. I'll stop trying to replicate it for now then, and will be patient for the next version.