dearwolf/LocalAI
1
0
Fork 0
mirror of https://github.com/mudler/LocalAI.git synced 2025-05-20 02:24:59 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 21

feat(p2p): allow to run multiple clusters in the same p2p network (#3128)

Browse source 
feat(p2p): allow to run multiple clusters in the same network

Allow to specify a network ID via CLI which allows to run multiple
clusters, logically separated within the same network (by using the same
shared token).

Note: This segregation is not "secure" by any means, anyone having the
network token can see the services available in all the network,
however, this provides a way to separate the inference endpoints.

This allows for instance to have a node which is both federated and
having attached a set of llama.cpp workers.

Signed-off-by: Ettore Di Giacinto <mudler@localai.io>
This commit is contained in:
Ettore Di Giacinto 2024-08-07 23:35:44 +02:00 • committed by GitHub
parent 2c8623dbb4
commit 36e185ba63
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
8 changed files with 50 additions and 26 deletions
Download patch file Download diff file Expand all files Collapse all files

9
core/cli/federated.go
View file

@ -8,14 +8,15 @@ import (
)
type FederatedCLI struct {
Address string `env:"LOCALAI_ADDRESS,ADDRESS" default:":8080" help:"Bind address for the API server" group:"api"`
Peer2PeerToken string `env:"LOCALAI_P2P_TOKEN,P2P_TOKEN,TOKEN" name:"p2ptoken" help:"Token for P2P mode (optional)" group:"p2p"`
LoadBalanced bool `env:"LOCALAI_LOAD_BALANCED,LOAD_BALANCED" default:"false" help:"Enable load balancing" group:"p2p"`
Address string `env:"LOCALAI_ADDRESS,ADDRESS" default:":8080" help:"Bind address for the API server" group:"api"`
Peer2PeerToken string `env:"LOCALAI_P2P_TOKEN,P2P_TOKEN,TOKEN" name:"p2ptoken" help:"Token for P2P mode (optional)" group:"p2p"`
LoadBalanced bool `env:"LOCALAI_LOAD_BALANCED,LOAD_BALANCED" default:"false" help:"Enable load balancing" group:"p2p"`
Peer2PeerNetworkID string `env:"LOCALAI_P2P_NETWORK_ID,P2P_NETWORK_ID" help:"Network ID for P2P mode, can be set arbitrarly by the user for grouping a set of instances." group:"p2p"`
}
func (f *FederatedCLI) Run(ctx *cliContext.Context) error {
fs := p2p.NewFederatedServer(f.Address, p2p.FederatedID, f.Peer2PeerToken, f.LoadBalanced)
fs := p2p.NewFederatedServer(f.Address, p2p.NetworkID(f.Peer2PeerNetworkID, p2p.FederatedID), f.Peer2PeerToken, f.LoadBalanced)
return fs.Start(context.Background())
}

11
core/cli/run.go
View file

@ -54,6 +54,7 @@ type RunCMD struct {
OpaqueErrors bool `env:"LOCALAI_OPAQUE_ERRORS" default:"false" help:"If true, all error responses are replaced with blank 500 errors. This is intended only for hardening against information leaks and is normally not recommended." group:"hardening"`
Peer2Peer bool `env:"LOCALAI_P2P,P2P" name:"p2p" default:"false" help:"Enable P2P mode" group:"p2p"`
Peer2PeerToken string `env:"LOCALAI_P2P_TOKEN,P2P_TOKEN,TOKEN" name:"p2ptoken" help:"Token for P2P mode (optional)" group:"p2p"`
Peer2PeerNetworkID string `env:"LOCALAI_P2P_NETWORK_ID,P2P_NETWORK_ID" help:"Network ID for P2P mode, can be set arbitrarly by the user for grouping a set of instances" group:"p2p"`
ParallelRequests bool `env:"LOCALAI_PARALLEL_REQUESTS,PARALLEL_REQUESTS" help:"Enable backends to handle multiple requests in parallel if they support it (e.g.: llama.cpp or vllm)" group:"backends"`
SingleActiveBackend bool `env:"LOCALAI_SINGLE_ACTIVE_BACKEND,SINGLE_ACTIVE_BACKEND" help:"Allow only one backend to be run at a time" group:"backends"`
PreloadBackendOnly bool `env:"LOCALAI_PRELOAD_BACKEND_ONLY,PRELOAD_BACKEND_ONLY" default:"false" help:"Do not launch the API services, only the preloaded models / backends are started (useful for multi-node setups)" group:"backends"`
@ -94,6 +95,7 @@ func (r *RunCMD) Run(ctx *cliContext.Context) error {
config.WithModelsURL(append(r.Models, r.ModelArgs...)...),
config.WithOpaqueErrors(r.OpaqueErrors),
config.WithEnforcedPredownloadScans(!r.DisablePredownloadScan),
config.WithP2PNetworkID(r.Peer2PeerNetworkID),
}
token := ""
@ -119,9 +121,9 @@ func (r *RunCMD) Run(ctx *cliContext.Context) error {
}
log.Info().Msg("Starting P2P server discovery...")
if err := p2p.ServiceDiscoverer(context.Background(), node, token, "", func(serviceID string, node p2p.NodeData) {
if err := p2p.ServiceDiscoverer(context.Background(), node, token, p2p.NetworkID(r.Peer2PeerNetworkID, ""), func(serviceID string, node p2p.NodeData) {
var tunnelAddresses []string
for _, v := range p2p.GetAvailableNodes("") {
for _, v := range p2p.GetAvailableNodes(p2p.NetworkID(r.Peer2PeerNetworkID, "")) {
if v.IsOnline() {
tunnelAddresses = append(tunnelAddresses, v.TunnelAddress)
} else {
@ -142,14 +144,15 @@ func (r *RunCMD) Run(ctx *cliContext.Context) error {
if err != nil {
return err
}
if err := p2p.ExposeService(context.Background(), "localhost", port, token, p2p.FederatedID); err != nil {
if err := p2p.ExposeService(context.Background(), "localhost", port, token, p2p.NetworkID(r.Peer2PeerNetworkID, p2p.FederatedID)); err != nil {
return err
}
node, err := p2p.NewNode(token)
if err != nil {
return err
}
if err := p2p.ServiceDiscoverer(context.Background(), node, token, p2p.FederatedID, nil); err != nil {
if err := p2p.ServiceDiscoverer(context.Background(), node, token, p2p.NetworkID(r.Peer2PeerNetworkID, p2p.FederatedID), nil); err != nil {
return err
}
}

17
core/cli/worker/worker_p2p.go
View file

@ -19,12 +19,13 @@ import (
)
type P2P struct {
WorkerFlags `embed:""`
Token string `env:"LOCALAI_TOKEN,LOCALAI_P2P_TOKEN,TOKEN" help:"P2P token to use"`
NoRunner bool `env:"LOCALAI_NO_RUNNER,NO_RUNNER" help:"Do not start the llama-cpp-rpc-server"`
RunnerAddress string `env:"LOCALAI_RUNNER_ADDRESS,RUNNER_ADDRESS" help:"Address of the llama-cpp-rpc-server"`
RunnerPort string `env:"LOCALAI_RUNNER_PORT,RUNNER_PORT" help:"Port of the llama-cpp-rpc-server"`
ExtraLLamaCPPArgs []string `env:"LOCALAI_EXTRA_LLAMA_CPP_ARGS,EXTRA_LLAMA_CPP_ARGS" help:"Extra arguments to pass to llama-cpp-rpc-server"`
WorkerFlags `embed:""`
Token string `env:"LOCALAI_TOKEN,LOCALAI_P2P_TOKEN,TOKEN" help:"P2P token to use"`
NoRunner bool `env:"LOCALAI_NO_RUNNER,NO_RUNNER" help:"Do not start the llama-cpp-rpc-server"`
RunnerAddress string `env:"LOCALAI_RUNNER_ADDRESS,RUNNER_ADDRESS" help:"Address of the llama-cpp-rpc-server"`
RunnerPort string `env:"LOCALAI_RUNNER_PORT,RUNNER_PORT" help:"Port of the llama-cpp-rpc-server"`
ExtraLLamaCPPArgs []string `env:"LOCALAI_EXTRA_LLAMA_CPP_ARGS,EXTRA_LLAMA_CPP_ARGS" help:"Extra arguments to pass to llama-cpp-rpc-server"`
Peer2PeerNetworkID string `env:"LOCALAI_P2P_NETWORK_ID,P2P_NETWORK_ID" help:"Network ID for P2P mode, can be set arbitrarly by the user for grouping a set of instances" group:"p2p"`
}
func (r *P2P) Run(ctx *cliContext.Context) error {
@ -59,7 +60,7 @@ func (r *P2P) Run(ctx *cliContext.Context) error {
p = r.RunnerPort
}
err = p2p.ExposeService(context.Background(), address, p, r.Token, "")
err = p2p.ExposeService(context.Background(), address, p, r.Token, p2p.NetworkID(r.Peer2PeerNetworkID, ""))
if err != nil {
return err
}
@ -99,7 +100,7 @@ func (r *P2P) Run(ctx *cliContext.Context) error {
}
}()
err = p2p.ExposeService(context.Background(), address, fmt.Sprint(port), r.Token, "")
err = p2p.ExposeService(context.Background(), address, fmt.Sprint(port), r.Token, p2p.NetworkID(r.Peer2PeerNetworkID, ""))
if err != nil {
return err
}