feat(#1107): hivectl agents restart / restart-all commands

This commit is contained in:
damocles 2026-06-03 10:45:48 +02:00 committed by mara
commit 26f2c1f59b
3 changed files with 116 additions and 5 deletions

View file

@ -3,11 +3,10 @@
//! Sibling binary to the `hive-c0re` daemon. Where `hive-c0re`'s
//! subcommands focus on the broker / approval / topology surface
//! (`spawn`, `kill`, `rebuild`, `approve` …), `hivectl` covers
//! host-side administration that doesn't need the daemon running —
//! starting with manual user provisioning on the bundled forge +
//! matrix containers when c0re's automatic boot-time sweep is
//! inappropriate (recovery, debugging, single-shot reprovisioning,
//! verifying the registration token path).
//! host-side administration — both manual provisioning operations
//! that work without the daemon running (forge, matrix, gateway) AND
//! daemon-assisted agent management (agents restart/restart-all) that
//! goes through the host admin socket.
//!
//! Verbs read configuration off the same on-disk paths c0re uses
//! (`/var/lib/hyperhive/forge-core-token`,
@ -64,6 +63,12 @@ enum Cmd {
#[command(subcommand)]
cmd: GatewayCmd,
},
/// Agent container management. Requires the hive-c0re daemon to be
/// running (connects to the host admin socket).
Agents {
#[command(subcommand)]
cmd: AgentsCmd,
},
}
#[derive(Subcommand)]
@ -196,6 +201,32 @@ enum GatewayCmd {
},
}
/// Default host admin socket path (same as `hive-c0re`'s default).
const DEFAULT_HOST_SOCKET: &str = "/run/hive/host.sock";
#[derive(Subcommand)]
enum AgentsCmd {
/// Stop and start a single agent container without rebuilding config.
/// Useful for "kick the container" when the process is stuck or the
/// container needs a clean restart without changing the NixOS config.
Restart {
/// Agent name (e.g. `damocles`, `ruth`).
name: String,
/// Path to the hive-c0re host admin socket.
#[arg(long, default_value = DEFAULT_HOST_SOCKET)]
socket: PathBuf,
},
/// Stop and restart ALL managed agent containers in sequence.
/// Iterates the live container list and restarts each one. Any per-agent
/// failure is reported at the end rather than stopping mid-run, so all
/// containers get a restart attempt.
RestartAll {
/// Path to the hive-c0re host admin socket.
#[arg(long, default_value = DEFAULT_HOST_SOCKET)]
socket: PathBuf,
},
}
#[tokio::main]
async fn main() -> Result<()> {
tracing_subscriber::fmt()
@ -230,6 +261,10 @@ async fn main() -> Result<()> {
GatewayCmd::DeleteUser { file, username } => gateway_delete_user(&file, &username),
GatewayCmd::ListUsers { file } => gateway_list_users(&file),
},
Cmd::Agents { cmd } => match cmd {
AgentsCmd::Restart { name, socket } => agents_restart(&socket, &name).await,
AgentsCmd::RestartAll { socket } => agents_restart_all(&socket).await,
},
}
}
@ -461,6 +496,48 @@ fn gateway_list_users(file: &Path) -> Result<()> {
Ok(())
}
// ---------------------------------------------------------------------------
// Agent management helpers (require daemon via host admin socket)
// ---------------------------------------------------------------------------
async fn agents_restart(socket: &Path, name: &str) -> Result<()> {
let resp = hive_c0re::client::request(socket, hive_sh4re::HostRequest::Restart {
name: name.to_owned(),
})
.await
.with_context(|| format!("connect to daemon socket {}", socket.display()))?;
if resp.ok {
println!("restarted: {name}");
Ok(())
} else {
bail!(
"restart {name}: {}",
resp.error.as_deref().unwrap_or("unknown error")
)
}
}
async fn agents_restart_all(socket: &Path) -> Result<()> {
let resp = hive_c0re::client::request(socket, hive_sh4re::HostRequest::RestartAll)
.await
.with_context(|| format!("connect to daemon socket {}", socket.display()))?;
let agents = resp.agents.as_deref().unwrap_or(&[]);
if agents.is_empty() {
println!("restart-all: no managed containers found");
} else {
for a in agents {
println!("restarted: {a}");
}
}
if !resp.ok {
bail!(
"restart-all: {}",
resp.error.as_deref().unwrap_or("unknown error")
);
}
Ok(())
}
/// Reject usernames containing `:` (field separator) or control chars
/// that would corrupt the htpasswd file format.
fn validate_htpasswd_username(username: &str) -> Result<()> {

View file

@ -140,6 +140,32 @@ async fn dispatch(req: &HostRequest, coord: Arc<Coordinator>) -> HostResponse {
});
HostResponse::success()
}
HostRequest::Restart { name } => {
tracing::info!(%name, "restart");
lifecycle::restart(name).await?;
HostResponse::success()
}
HostRequest::RestartAll => {
tracing::info!("restart-all");
let agents = lifecycle::list().await?;
let mut errors: Vec<String> = Vec::new();
for agent in &agents {
if let Err(e) = lifecycle::restart(agent).await {
tracing::warn!(%agent, error = ?e, "restart-all: failed to restart agent");
errors.push(format!("{agent}: {e:#}"));
}
}
if errors.is_empty() {
HostResponse::list(agents)
} else {
HostResponse {
ok: false,
error: Some(errors.join("; ")),
agents: Some(agents),
approvals: None,
}
}
}
HostRequest::Destroy { name, purge } => {
actions::destroy(&coord, name, *purge).await?;
HostResponse::success()

View file

@ -31,6 +31,14 @@ pub enum HostRequest {
#[serde(default)]
purge: bool,
},
/// Stop and start a managed container without rebuilding config.
/// For "kick the container" operations that don't touch the flake or
/// nspawn flags. Mirrors `lifecycle::restart` (kill + start).
Restart { name: String },
/// Stop and restart all managed containers in sequence. Convenience
/// wrapper for `hivectl agents restart-all`; iterates the live
/// container list and restarts each one.
RestartAll,
/// Apply pending config to a managed container.
Rebuild { name: String },
/// List managed containers.