LXC install support?

[hemna]

I would like to run this in my homelab as a proxmox lxc container instead of docker. I don't see any docs on installing this outside of docker?

[BeryJu]

You can in theory just run the binary out of the container on an LXC container (or any linux machine); the binary is not currently published separately though

[kri164]

Try something as

podman pull ghcr.io/beryju/gravity:v0.27.1
podman run --name gravity -d ghcr.io/beryju/gravity:v0.27.1
podman cp gravity:/usr/bin/gravity /tmp/gravity-0.27.1.bin
podman stop gravity
podman rm gravity
podman image rm gravity:v0.27.1

scp /tmp/gravity-0.27.1.bin [email protected]:/usr/bin/gravity

[hemna]

ok I tried that and it failed.

root@gravity1:~# podman image rm gravity:v0.27.1
Untagged: ghcr.io/beryju/gravity:v0.27.1
Deleted: cfeb1677fc20f6a12d2476fbdee278e34a8f46f65f819e72764605b9d310200f
root@gravity1:~# scp /tmp/gravity-0.27.1.bin /usr/bin/gravity
root@gravity1:~# gravity --help
Start gravity instance

Usage:
  gravity [command]

Available Commands:
  cli         Interact with a running Gravity server
  completion  Generate the autocompletion script for the specified shell
  help        Help about any command
  server      Run Gravity server

Flags:
  -h, --help      help for gravity
  -v, --version   version for gravity

Use "gravity [command] --help" for more information about a command.
root@gravity1:~# gravity server
{"level":"info","ts":1758288719.9295077,"logger":"instance","msg":"Gravity starting","instance":"gravity1","version":"0.27.1","instance":"gravity1","version":"0.27.1"}
{"level":"info","ts":1758288719.9295683,"logger":"instance","msg":"'etcd' in bootstrap roles, starting embedded etcd","instance":"gravity1","version":"0.27.1","instance":"gravity1"}
...
...
{"level":"warn","ts":1758288721.259253,"logger":"role.etcd","msg":"request stats","instance":"gravity1","version":"0.27.1","start time":1758288721.2575808,"time spent":1,"remote":"127.0.0.1:36086","response type":"/etcdserverpb.KV/DeleteRange","request count":0,"request size":69,"response count":1,"response size":29,"request content":"key:\"/gravity/discovery/subnets/instance-subnet-gravity1/bb999622cb05e0e\" "}
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x10 pc=0x13531ba]

goroutine 270 [running]:
github.com/Ullaakut/nmap/v2.(*Scanner).Args(...)
        /go/pkg/mod/github.com/!ullaakut/nmap/[email protected]/nmap.go:1622
beryju.io/gravity/pkg/roles/discovery.(*Subnet).RunDiscovery(0xc000ce44d0, {0x6aa47a0?, 0x993ab60?})
        /workspace/pkg/roles/discovery/subnet.go:103 +0x97a
created by beryju.io/gravity/pkg/roles/discovery.New.func3 in goroutine 1
        /workspace/pkg/roles/discovery/role.go:83 +0x3b1
root@gravity1:~#

[BeryJu]

587f032 should fix this issue for the next release

[BeryJu]

Also this error most likely is happening due to nmap not being installed in the LXC container

[hemna]

I have resorted to running docker in an lxc container and that's working.