Skip to content

server: Fix router proxying to child processes when --host is specified #18054

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
wbtek wants to merge 1 commit into from
Closed

server: Fix router proxying to child processes when --host is specified #18054

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
17 changes: 11 additions & 6 deletions tools/server/server-models.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -209,7 +209,7 @@ void server_presets::render_args(server_model_meta & meta) {
}
// 3. control args (from router)
// set control values
preset.options[control_args["LLAMA_ARG_HOST"]] = CHILD_ADDR;
preset.options[control_args["LLAMA_ARG_HOST"]] = meta.hostname;
preset.options[control_args["LLAMA_ARG_PORT"]] = std::to_string(meta.port);
preset.options[control_args["LLAMA_ARG_ALIAS"]] = meta.name;
if (meta.in_cache) {
Expand Down Expand Up @@ -308,7 +308,8 @@ void server_models::load_models() {
/* status */ SERVER_MODEL_STATUS_UNLOADED,
/* last_used */ 0,
/* args */ std::vector<std::string>(),
/* exit_code */ 0
/* exit_code */ 0,
/* host_name */ CHILD_ADDR
};
add_model(std::move(meta));
}
Expand All @@ -330,7 +331,8 @@ void server_models::load_models() {
/* status */ SERVER_MODEL_STATUS_UNLOADED,
/* last_used */ 0,
/* args */ std::vector<std::string>(),
/* exit_code */ 0
/* exit_code */ 0,
/* host_name */ CHILD_ADDR
};
add_model(std::move(meta));
}
Expand All @@ -348,7 +350,8 @@ void server_models::load_models() {
/* status */ SERVER_MODEL_STATUS_UNLOADED,
/* last_used */ 0,
/* args */ std::vector<std::string>(),
/* exit_code */ 0
/* exit_code */ 0,
/* host_name */ CHILD_ADDR
};
add_model(std::move(meta));
}
Expand Down Expand Up @@ -510,6 +513,7 @@ void server_models::load(const std::string & name) {
inst.meta.port = get_free_port();
inst.meta.status = SERVER_MODEL_STATUS_LOADING;
inst.meta.last_used = ggml_time_ms();
inst.meta.hostname = base_params.hostname; // Use router's cmdline value
Copy link
Copy Markdown
Contributor

@aldehir aldehir Dec 15, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

0.0.0.0 does not route to 127.0.0.1 on Windows like it does on Linux.

A quick fix may be to only change to 127.0.0.1 on --host 0.0.0.0, otherwise use the provided IP.

#17962 (comment)


if (inst.meta.port <= 0) {
throw std::runtime_error("failed to get a port number");
Expand Down Expand Up @@ -698,7 +702,7 @@ server_http_res_ptr server_models::proxy_request(const server_http_req & req, co
SRV_INF("proxying request to model %s on port %d\n", name.c_str(), meta->port);
auto proxy = std::make_unique<server_http_proxy>(
method,
CHILD_ADDR,
base_params.hostname,
Copy link
Copy Markdown
Contributor

@aldehir aldehir Dec 15, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Same thing here.

meta->port,
req.path,
req.headers,
Expand Down Expand Up @@ -726,7 +730,8 @@ std::thread server_models::setup_child_server(const common_params & base_params,
body["value"] = server_model_status_to_string(SERVER_MODEL_STATUS_LOADED);
req.body = body.dump();

SRV_INF("notifying router server (port=%d) that model %s is ready\n", router_port, name.c_str());
SRV_INF("notifying router server (%s:port=%d) that model %s is ready\n",
base_params.hostname.c_str(), router_port, name.c_str());
auto result = cli.send(std::move(req));
if (result.error() != httplib::Error::Success) {
auto err_str = httplib::to_string(result.error());
Expand Down
1 change: 1 addition & 0 deletions tools/server/server-models.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -58,6 +58,7 @@ struct server_model_meta {
int64_t last_used = 0; // for LRU unloading
std::vector<std::string> args; // args passed to the model instance, will be populated by render_args()
int exit_code = 0; // exit code of the model instance process (only valid if status == FAILED)
std::string hostname;

bool is_active() const {
return status == SERVER_MODEL_STATUS_LOADED || status == SERVER_MODEL_STATUS_LOADING;
Expand Down