refactor(packaging): rely on gateway runtime defaults (#1415)

* fix(packaging): use gateway TOML config in packages

* refactor(packaging): rely on gateway runtime defaults

Author: drew

crates/openshell-cli/src/run.rs

@@ -675,8 +675,8 @@ fn is_loopback_gateway_endpoint(endpoint: &str) -> bool {
 /// would serve this endpoint.
 ///
 /// Loopback endpoints (`localhost`, `127.0.0.1`, `::1`) resolve to the
-/// `"openshell"` gateway name, matching the convention used by
-/// `init-pki.sh` and the TLS cert resolver in `tls.rs`.
+/// `"openshell"` gateway name, matching the convention used by local
+/// `openshell-gateway generate-certs` and the TLS cert resolver in `tls.rs`.
 fn mtls_certs_exist_for_endpoint(name: &str, endpoint: &str) -> bool {
     let cert_name = if is_loopback_gateway_endpoint(endpoint) {
         "openshell"
@@ -901,7 +901,7 @@ pub async fn gateway_add(
 
     // Derive a gateway name from the hostname when none is provided.
     // Loopback endpoints use the canonical "openshell" name, matching the
-    // convention in init-pki.sh and default_tls_dir.
+    // convention in local cert generation and default_tls_dir.
     let derived_name;
     let name = if let Some(n) = name {
         n
@@ -7240,7 +7240,7 @@ mod tests {
             });
 
             // Loopback endpoints derive the canonical "openshell" gateway
-            // name, matching init-pki.sh and default_tls_dir conventions.
+            // name, matching local cert generation and default_tls_dir conventions.
             let metadata = load_gateway_metadata("openshell").expect("load stored gateway");
             assert_eq!(metadata.auth_mode.as_deref(), Some("plaintext"));
             assert!(!metadata.is_remote);

crates/openshell-core/src/config.rs

@@ -22,7 +22,7 @@ use std::str::FromStr;
 pub const DEFAULT_SSH_PORT: u16 = 2222;
 
 /// Default gateway server port.
-pub const DEFAULT_SERVER_PORT: u16 = 8080;
+pub const DEFAULT_SERVER_PORT: u16 = 17670;
 
 /// Default container stop timeout in seconds (SIGTERM → SIGKILL).
 pub const DEFAULT_STOP_TIMEOUT_SECS: u32 = 10;
@@ -34,7 +34,7 @@ pub const DEFAULT_DOCKER_NETWORK_NAME: &str = "openshell-docker";
 pub const DEFAULT_SERVICE_ROUTING_DOMAIN: &str = "openshell.localhost";
 
 /// Default OCI image for the openshell-sandbox supervisor binary.
-pub const DEFAULT_SUPERVISOR_IMAGE: &str = "openshell/supervisor:latest";
+pub const DEFAULT_SUPERVISOR_IMAGE: &str = "ghcr.io/nvidia/openshell/supervisor:latest";
 
 /// CDI device identifier for requesting all NVIDIA GPUs.
 pub const CDI_GPU_DEVICE_ALL: &str = "nvidia.com/gpu=all";
@@ -451,7 +451,7 @@ impl Default for ServiceRoutingConfig {
 }
 
 fn default_bind_address() -> SocketAddr {
-    "127.0.0.1:8080".parse().expect("valid default address")
+    "127.0.0.1:17670".parse().expect("valid default address")
 }
 
 fn default_service_routing_domains() -> Vec<String> {
@@ -557,7 +557,7 @@ mod tests {
 
     #[test]
     fn config_defaults_to_loopback_bind_address() {
-        let expected: SocketAddr = "127.0.0.1:8080".parse().expect("valid address");
+        let expected: SocketAddr = "127.0.0.1:17670".parse().expect("valid address");
         assert_eq!(Config::new(None).bind_address, expected);
     }
 

crates/openshell-core/src/paths.rs

@@ -29,6 +29,24 @@ pub fn openshell_config_dir() -> Result<PathBuf> {
     Ok(xdg_config_dir()?.join("openshell"))
 }
 
+/// Resolve the XDG state base directory.
+///
+/// Returns `$XDG_STATE_HOME` if set, otherwise `$HOME/.local/state`.
+pub fn xdg_state_dir() -> Result<PathBuf> {
+    if let Ok(path) = std::env::var("XDG_STATE_HOME") {
+        return Ok(PathBuf::from(path));
+    }
+    let home = std::env::var("HOME")
+        .into_diagnostic()
+        .wrap_err("HOME is not set")?;
+    Ok(PathBuf::from(home).join(".local").join("state"))
+}
+
+/// The top-level `OpenShell` state directory: `$XDG_STATE_HOME/openshell/`.
+pub fn openshell_state_dir() -> Result<PathBuf> {
+    Ok(xdg_state_dir()?.join("openshell"))
+}
+
 /// Resolve the XDG data base directory.
 ///
 /// Returns `$XDG_DATA_HOME` if set, otherwise `$HOME/.local/share`.
@@ -130,6 +148,15 @@ mod tests {
         );
     }
 
+    #[test]
+    fn openshell_state_dir_appends_openshell() {
+        let dir = openshell_state_dir().unwrap();
+        assert!(
+            dir.ends_with("openshell"),
+            "expected path ending with 'openshell', got: {dir:?}"
+        );
+    }
+
     #[cfg(unix)]
     #[test]
     fn create_dir_restricted_sets_0o700() {

crates/openshell-driver-docker/src/lib.rs

@@ -252,12 +252,6 @@ impl DockerComputeDriver {
         docker_config: &DockerComputeConfig,
         supervisor_readiness: Arc<dyn SupervisorReadiness>,
     ) -> CoreResult<Self> {
-        if docker_config.grpc_endpoint.trim().is_empty() {
-            return Err(Error::config(
-                "grpc_endpoint is required when using the docker compute driver",
-            ));
-        }
-
         let docker = Docker::connect_with_local_defaults()
             .map_err(|err| Error::execution(format!("failed to create Docker client: {err}")))?;
         let version = docker.version().await.map_err(|err| {
@@ -281,14 +275,24 @@ impl DockerComputeDriver {
         let host_gateway_ip = parse_optional_host_gateway_ip(&docker_config.host_gateway_ip)?;
         let gateway_route =
             docker_gateway_route(&info, bridge_gateway_ip, gateway_port, host_gateway_ip);
+        let mut docker_config = docker_config.clone();
+        if docker_config.grpc_endpoint.trim().is_empty() {
+            let scheme = if docker_guest_tls_configured(&docker_config) {
+                "https"
+            } else {
+                "http"
+            };
+            docker_config.grpc_endpoint =
+                format!("{scheme}://{HOST_OPENSHELL_INTERNAL}:{gateway_port}");
+        }
         let grpc_endpoint = docker_container_openshell_endpoint(
             &docker_config.grpc_endpoint,
             HOST_OPENSHELL_INTERNAL,
             gateway_port,
         );
         let daemon_arch = normalize_docker_arch(version.arch.as_deref().unwrap_or_default());
-        let supervisor_bin = resolve_supervisor_bin(&docker, docker_config, &daemon_arch).await?;
-        let guest_tls = docker_guest_tls_paths(docker_config)?;
+        let supervisor_bin = resolve_supervisor_bin(&docker, &docker_config, &daemon_arch).await?;
+        let guest_tls = docker_guest_tls_paths(&docker_config)?;
 
         let driver = Self {
             docker: Arc::new(docker),
@@ -2009,6 +2013,12 @@ pub(crate) fn validate_linux_elf_binary(path: &Path) -> CoreResult<()> {
     Ok(())
 }
 
+fn docker_guest_tls_configured(docker_config: &DockerComputeConfig) -> bool {
+    docker_config.guest_tls_ca.is_some()
+        && docker_config.guest_tls_cert.is_some()
+        && docker_config.guest_tls_key.is_some()
+}
+
 pub(crate) fn docker_guest_tls_paths(
     docker_config: &DockerComputeConfig,
 ) -> CoreResult<Option<DockerGuestTlsPaths>> {

crates/openshell-driver-docker/src/tests.rs

@@ -74,23 +74,23 @@ fn container_visible_endpoint_rewrites_loopback_hosts() {
             HOST_OPENSHELL_INTERNAL,
             DEFAULT_SERVER_PORT,
         ),
-        "https://host.openshell.internal:8080/"
+        "https://host.openshell.internal:17670/"
     );
     assert_eq!(
         docker_container_openshell_endpoint(
             "http://127.0.0.1:8080",
             HOST_OPENSHELL_INTERNAL,
             DEFAULT_SERVER_PORT,
         ),
-        "http://host.openshell.internal:8080/"
+        "http://host.openshell.internal:17670/"
     );
     assert_eq!(
         docker_container_openshell_endpoint(
             "https://gateway.internal:8443",
             HOST_OPENSHELL_INTERNAL,
             DEFAULT_SERVER_PORT,
         ),
-        "https://host.openshell.internal:8080/"
+        "https://host.openshell.internal:17670/"
     );
 }
 
@@ -273,7 +273,7 @@ fn docker_gateway_route_uses_bridge_gateway_for_linux_docker() {
     assert_eq!(
         route,
         DockerGatewayRoute::Bridge {
-            bind_address: "172.18.0.1:8080".parse().unwrap(),
+            bind_address: "172.18.0.1:17670".parse().unwrap(),
             host_alias_ip: IpAddr::V4(Ipv4Addr::new(172, 18, 0, 1)),
         }
     );
@@ -303,7 +303,7 @@ fn docker_gateway_route_prefers_configured_host_gateway_ip() {
     assert_eq!(
         route,
         DockerGatewayRoute::Bridge {
-            bind_address: "172.20.0.4:8080".parse().unwrap(),
+            bind_address: "172.20.0.4:17670".parse().unwrap(),
             host_alias_ip: IpAddr::V4(Ipv4Addr::new(172, 20, 0, 4)),
         }
     );

crates/openshell-driver-podman/NETWORKING.md

@@ -357,9 +357,9 @@ Supervisor proxy in container netns
 
 The Podman driver auto-detects the callback endpoint scheme based on whether
 TLS client certificates are configured. When the RPM's auto-generated PKI is in
-place, the endpoint is `https://host.containers.internal:8080` and the
+place, the endpoint is `https://host.containers.internal:17670` and the
 supervisor connects with mTLS. Without TLS configuration, it falls back to
-`http://host.containers.internal:8080`.
+`http://host.containers.internal:<gateway-port>`.
 
 ```text
 Supervisor in container netns
@@ -382,10 +382,9 @@ Gateway
   9. Same gRPC channel reused for RelayStream calls
 ```
 
-The gateway binds to `0.0.0.0` by default in the RPM packaging. mTLS prevents
-unauthenticated access even though the gateway is reachable from the network.
-Client certificates are auto-generated by `init-pki.sh` on first start and
-bind-mounted into sandbox containers by the Podman driver.
+The gateway binds to `127.0.0.1:17670` by default in the RPM packaging. Client
+certificates are auto-generated by `openshell-gateway generate-certs` on first
+start and bind-mounted into sandbox containers by the Podman driver.
 
 ## Differences from the Kubernetes Driver
 
@@ -412,7 +411,7 @@ published ports, or the supervisor relay.
 
 | Port | Component | Purpose |
 |---|---|---|
-| `8080` | Gateway | gRPC and HTTP multiplexed default server port. |
+| `17670` | Gateway | Default local gRPC and HTTP multiplexed server port. |
 | `2222` | Sandbox | Container port mapping default for the SSH compatibility port. |
 | `3128` | Sandbox proxy | HTTP CONNECT proxy inside the sandbox network model. |
 | `0` | Host | Ephemeral host port requested for the container SSH compatibility port. |

crates/openshell-driver-podman/README.md

@@ -120,10 +120,10 @@ connection back to the gateway. On SELinux systems, the bind mounts include
 Podman's shared relabel option so the container process can read the files.
 
 The RPM packaging auto-generates a self-signed PKI on first start via
-`init-pki.sh`. Client certs are placed in the CLI auto-discovery directory
-(`~/.config/openshell/gateways/openshell/mtls/`) so the CLI connects with mTLS
-without manual configuration. See `deploy/rpm/CONFIGURATION.md` for the full
-RPM configuration reference.
+`openshell-gateway generate-certs`. Client certs are placed in the CLI
+auto-discovery directory (`~/.config/openshell/gateways/openshell/mtls/`) so
+the CLI connects with mTLS without manual configuration. See
+`deploy/rpm/CONFIGURATION.md` for the full RPM configuration reference.
 
 ## Network Model
 
@@ -134,7 +134,7 @@ the supervisor for sandbox process isolation.
 ```mermaid
 graph TB
     subgraph Host
-        GW["Gateway Server<br/>127.0.0.1:8080"]
+        GW["Gateway Server<br/>127.0.0.1:17670"]
         PS["Podman Socket"]
     end
 
@@ -289,11 +289,11 @@ Podman resources after out-of-band container removal or label drift.
 | `OPENSHELL_SANDBOX_IMAGE` | `--sandbox-image` | From gateway config | Default OCI image for sandboxes. |
 | `OPENSHELL_SANDBOX_IMAGE_PULL_POLICY` | `--sandbox-image-pull-policy` | `missing` | Pull policy: `always`, `missing`, `never`, or `newer`. |
 | `OPENSHELL_GRPC_ENDPOINT` | `--grpc-endpoint` | Auto-detected via `host.containers.internal` | Gateway gRPC endpoint for sandbox callbacks. |
-| `OPENSHELL_GATEWAY_PORT` | `--gateway-port` | `8080` | Gateway port used for endpoint auto-detection by the standalone binary. |
+| `OPENSHELL_GATEWAY_PORT` | `--gateway-port` | `17670` | Gateway port used for endpoint auto-detection by the standalone binary. |
 | `OPENSHELL_NETWORK_NAME` | `--network-name` | `openshell` | Podman bridge network name. |
 | `OPENSHELL_SANDBOX_SSH_SOCKET_PATH` | `--sandbox-ssh-socket-path` | `/run/openshell/ssh.sock` | Supervisor Unix socket path in `PodmanComputeConfig`. |
 | `OPENSHELL_STOP_TIMEOUT` | `--stop-timeout` | `10` | Container stop timeout in seconds. |
-| `OPENSHELL_SUPERVISOR_IMAGE` | `--supervisor-image` | `openshell/supervisor:latest` through the gateway, required standalone | OCI image containing the supervisor binary. |
+| `OPENSHELL_SUPERVISOR_IMAGE` | `--supervisor-image` | `ghcr.io/nvidia/openshell/supervisor:latest` through the gateway, required standalone | OCI image containing the supervisor binary. |
 | `OPENSHELL_PODMAN_TLS_CA` | `--podman-tls-ca` | unset | Host path to the CA certificate mounted for sandbox mTLS. |
 | `OPENSHELL_PODMAN_TLS_CERT` | `--podman-tls-cert` | unset | Host path to the client certificate mounted for sandbox mTLS. |
 | `OPENSHELL_PODMAN_TLS_KEY` | `--podman-tls-key` | unset | Host path to the client private key mounted for sandbox mTLS. |

crates/openshell-driver-podman/src/container.rs

@@ -1036,7 +1036,7 @@ mod tests {
         let vol = &image_volumes[0];
         assert_eq!(
             vol["source"].as_str(),
-            Some("openshell/supervisor:latest"),
+            Some("ghcr.io/nvidia/openshell/supervisor:latest"),
             "image volume source should be the supervisor image"
         );
         assert_eq!(

crates/openshell-server/src/certgen.rs

@@ -8,9 +8,9 @@
 //! - **Kubernetes mode** (default): create two `kubernetes.io/tls` Secrets
 //!   in the supplied namespace. Used by the Helm pre-install hook. Requires
 //!   `--namespace`, `--server-secret-name`, `--client-secret-name`.
-//! - **Local mode** (`--output-dir <DIR>`): write PEMs to a filesystem layout
-//!   used by the RPM systemd unit's `ExecStartPre`. Also copies client
-//!   materials to
+//! - **Local mode** (`--output-dir <DIR>`): write PEMs to the local package
+//!   filesystem layout. Used by systemd units' `ExecStartPre`. Also copies
+//!   client materials to
 //!   `$XDG_CONFIG_HOME/openshell/gateways/openshell/mtls/` so the local CLI
 //!   picks them up automatically.
 //!