v0.4.67: NFSv3 alongside SMB (in-process via nfs3_client crate)

NFS is back — done right this time. v0.4.67 ships a pure-Rust NFSv3
client (`nfs3_client` 0.9 from the xetdata/Vaiz crate family) running
in-process inside the openpxe binary. No `mount.nfs`, no kernel
modules, no `CAP_SYS_ADMIN`, no subprocess. Works in every container
that the v0.4.65 SMB path works in (Unraid included).

The v0.4.65 SMB path stays as-is. Operators get both protocols
side-by-side and pick whichever their NAS prefers — or use both
together. NFSv3 has one architectural advantage over the SMB
userspace path: HTTP Range requests work for NFS-sourced ISOs
because NFSv3 READ3 takes an explicit offset. SMB-sourced ISOs still
return 416 for ranges (smbclient CLI can't seek mid-stream).

## What's new

- `crates/iso-store/src/nfs_share.rs` — `NfsShareManager` mirroring
  `SmbShareManager` structurally. Lists ISOs via READDIR3+LOOKUP3+
  GETATTR3, streams files via READ3 in 64 KiB chunks piped to axum
  body streams. Uses `connect_from_privileged_port(false)` because
  the openpxe binary runs as uid 10001 — most modern NFS servers
  allow that; a server that demands privileged ports needs
  `insecure` in /etc/exports, and the hint translation calls that
  out specifically.
- `IsoSource::Nfs { share_id, relative_path }` variant alongside the
  existing `Smb`. `IsoStore::iso_path_for` returns None for both;
  the HTTP handler dispatches to the right share manager.
- `/api/nfs-shares` CRUD + scan endpoints, parallel to
  `/api/smb-shares`. `POST` body: `{ server, export, port? }`.
- `nfs` terminal command back (this time as in-process, not kernel
  mount): `list | add <srv>:<export> [port] | remove | scan`. The
  v0.4.64 `nfs` command name pointing at kernel mount is moot
  history — same name, completely different mechanism.
- Storage tab: a new NFS shares card sits directly below the SMB
  shares card. The form is simpler (no auth fields) since NFSv3
  uses AUTH_SYS and access is gated server-side by client IP.
- Dashboard "Images available" tile sums SMB + NFS reachable shares
  into a generic "N remote shares" line.

## What's the same

- The structured `{error, stderr, hint}` JSON shape on failures
  matches the SMB API exactly, so the UI's error banner renders
  identically.
- Hint translation: NFS3ERR_ACCES → "exports list", NFS3ERR_NOENT →
  "export path doesn't exist", `mount denied` → "/etc/exports may
  need `insecure`", timeouts → "check IP/port/firewall".
- Persistence: `<work_dir>/nfs_shares.json`. No conflict with the
  long-dead v0.4.64 `nfs.json`.

## Why nfs3_client

User picked it: pure-Rust matches the architecture, NFSv3 covers the
real-world cases, AUTH_SYS keeps the UI simple. The crate is at
0.9.0, MIT/Unlicense, rust-version 1.88 (we're on 1.95). Tokio
feature flag enabled. Image size unchanged at compile time — single
musl static binary, no extra OS packages.

## Tests

160 passing (was 150 in v0.4.66, +10):
- nfs_share parser: stable share ids, server normalization (smb://,
  cifs://, \\, // all stripped).
- hint_for(): NFS3ERR_ACCES, NFS3ERR_NOENT, mount denied, unknown.
- status_label() covers the common nfsstat3 codes.
- HTTP integration: nfs-shares list starts empty, missing server
  rejected, export without leading slash rejected.

`cargo clippy --workspace --all-targets -- -D warnings` clean.

Co-Authored-By: Claude Opus 4.7 (1M context) <[email protected]>
This commit is contained in:
Miles Ward
2026-05-28 12:56:46 -04:00
co-authored by Claude Opus 4.7
parent 9f66c269c4
commit 3f9d8568f0
12 changed files with 1466 additions and 69 deletions
+141 -14
View File
@@ -35,7 +35,7 @@ use openpxe_core::{
MAX_LOGO_BYTES,
};
use openpxe_ipxe_assets::asset_bytes;
use openpxe_iso_store::{IsoCategory, IsoMeta, IsoSource, SmbAddRequest};
use openpxe_iso_store::{IsoCategory, IsoMeta, IsoSource, NfsAddRequest, SmbAddRequest};
use serde::Deserialize;
use serde_json::json;
use std::net::SocketAddr;
@@ -141,6 +141,12 @@ pub fn build_router(state: AppState) -> Router {
.route("/api/smb-shares", get(api_smb_shares_list).post(api_smb_shares_add))
.route("/api/smb-shares/:id", delete(api_smb_shares_remove))
.route("/api/smb-shares/:id/scan", post(api_smb_shares_scan))
// v0.4.67: NFSv3 share manager (pure-Rust in-process client).
// Ships alongside SMB. Routes are parallel so the UI can
// reuse the same form/error/hint rendering for both.
.route("/api/nfs-shares", get(api_nfs_shares_list).post(api_nfs_shares_add))
.route("/api/nfs-shares/:id", delete(api_nfs_shares_remove))
.route("/api/nfs-shares/:id/scan", post(api_nfs_shares_scan))
// Phase 4: Network info (read-only) + DNS edit.
.route("/api/network", get(api_network).put(api_network_put))
// Phase 4: live-log stream + recent buffer for the Terminal tab.
@@ -696,6 +702,58 @@ async fn iso_raw(
Err(e) => (StatusCode::BAD_GATEWAY, format!("smb stream: {e}")).into_response(),
}
}
IsoSource::Nfs {
share_id,
relative_path,
} => {
// v0.4.67: NFS sources support Range requests because
// NFSv3 READ3 takes an explicit offset. We resolve the
// requested byte range here and pass start/len down to
// the streamer which seeks into the file via READ3.
let total = meta.size_bytes;
let range = match parse_range(headers.get(header::RANGE), total) {
Some(triple) => triple,
None if headers.get(header::RANGE).is_some() => {
return Response::builder()
.status(StatusCode::RANGE_NOT_SATISFIABLE)
.header(header::CONTENT_RANGE, format!("bytes */{total}"))
.body(Body::empty())
.unwrap();
}
// No Range header — serve the whole file.
None => (0, total.saturating_sub(1), false),
};
let (start, end, partial) = range;
let len = if total == 0 { 0 } else { end - start + 1 };
let max_len = if total == 0 { None } else { Some(len) };
match state
.nfs_shares
.stream_iso(share_id, relative_path, start, max_len)
.await
{
Ok(stream) => {
let body = Body::from_stream(stream);
let status = if partial {
StatusCode::PARTIAL_CONTENT
} else {
StatusCode::OK
};
let mut builder = Response::builder()
.status(status)
.header(header::CONTENT_TYPE, "application/octet-stream")
.header(header::ACCEPT_RANGES, "bytes")
.header(header::CONTENT_LENGTH, len);
if partial {
builder = builder.header(
header::CONTENT_RANGE,
format!("bytes {start}-{end}/{total}"),
);
}
builder.body(body).unwrap()
}
Err(e) => (StatusCode::BAD_GATEWAY, format!("nfs stream: {e}")).into_response(),
}
}
}
}
@@ -1095,6 +1153,19 @@ async fn api_docs() -> Json<serde_json::Value> {
"summary": "Re-list a share for new ISOs."},
],
},
{
"name": "NFS shares",
"endpoints": [
{"method": "GET", "path": "/api/nfs-shares",
"summary": "List configured NFSv3 shares with connection state and iso counts."},
{"method": "POST", "path": "/api/nfs-shares",
"summary": "Register an NFSv3 share. Body: { server, export, port? }. Auth is AUTH_SYS only; access control is by client IP on the server side."},
{"method": "DELETE", "path": "/api/nfs-shares/:id",
"summary": "Forget a share and drop its entries from the ISO store."},
{"method": "POST", "path": "/api/nfs-shares/:id/scan",
"summary": "Re-list a share for new ISOs."},
],
},
{
"name": "Network",
"endpoints": [
@@ -1509,11 +1580,13 @@ async fn api_list_clients(State(state): State<AppState>) -> Json<serde_json::Val
async fn api_status(State(state): State<AppState>) -> Json<serde_json::Value> {
let smb = state.smb.as_ref().map(|s| s.snapshot());
// v0.4.65: NFS replaced with SMB shares. The dashboard metric
// shape stays similar (count + reachable) so the UI doesn't have
// to change its top-line tiles.
// v0.4.65+v0.4.67: external storage shares — SMB (userspace
// smbclient) and NFS (in-process nfs3_client). Dashboard tile
// sums both so operators see a single "reachable shares" number.
let smb_shares = state.smb_shares.list();
let smb_reachable = smb_shares.iter().filter(|m| m.reachable).count();
let nfs_shares = state.nfs_shares.list();
let nfs_reachable = nfs_shares.iter().filter(|m| m.reachable).count();
let isos = state.iso_store.list();
let clients = state.clients.list();
let queue_entries = state.queue.list();
@@ -1532,7 +1605,9 @@ async fn api_status(State(state): State<AppState>) -> Json<serde_json::Value> {
state
.metrics
.set_queue_counts(queue_entries.len() as u64, imaging as u64);
state.metrics.set_nfs_active(smb_reachable as u64);
state
.metrics
.set_nfs_active((smb_reachable + nfs_reachable) as u64);
state.metrics.record_http(openpxe_core::HttpRoute::Api);
let now = time::OffsetDateTime::now_utc();
let uptime_secs = (now - state.started_at).whole_seconds().max(0);
@@ -1547,12 +1622,13 @@ async fn api_status(State(state): State<AppState>) -> Json<serde_json::Value> {
"ipxe_assets": openpxe_ipxe_assets::list_assets(),
"settings": state.settings.snapshot(),
"smb": smb,
// Keep the field names for now so existing UI bindings on
// `iso_count2` / `client_count2` / sidebar counters keep
// working. They cover "external storage shares" generically;
// v0.4.65 the source is SMB instead of NFS.
"smb_share_count": smb_shares.len(),
"smb_share_reachable": smb_reachable,
// v0.4.67: NFSv3 share counts. The dashboard tile sums these
// with the SMB counts above ("N shares reachable") so the
// top-line metric works regardless of protocol mix.
"nfs_share_count": nfs_shares.len(),
"nfs_share_reachable": nfs_reachable,
"host_bindings": state.hosts.len(),
"custom_logo": state.branding.has_logo(),
"uptime_secs": uptime_secs,
@@ -1808,6 +1884,47 @@ async fn api_smb_shares_scan(
}
}
// ─── NFS share API (v0.4.67) ───────────────────────────────────────────────
//
// Parallel to the SMB shares API. The pure-Rust NFSv3 client
// (`nfs3_client`) gives us in-process listing and streaming, no
// subprocess. Unlike SMB, NFS-sourced ISOs support HTTP Range
// requests — NFSv3 READ3 takes an explicit offset.
async fn api_nfs_shares_list(State(state): State<AppState>) -> Json<serde_json::Value> {
Json(json!({ "shares": state.nfs_shares.list() }))
}
async fn api_nfs_shares_add(
State(state): State<AppState>,
Json(req): Json<NfsAddRequest>,
) -> Response {
match state.nfs_shares.add(req).await {
Ok(s) => (StatusCode::CREATED, Json(s)).into_response(),
Err(err) => (StatusCode::BAD_REQUEST, Json(err)).into_response(),
}
}
async fn api_nfs_shares_remove(
State(state): State<AppState>,
AxumPath(id): AxumPath<String>,
) -> Response {
match state.nfs_shares.remove(&id).await {
Ok(()) => StatusCode::NO_CONTENT.into_response(),
Err(e) => (StatusCode::INTERNAL_SERVER_ERROR, format!("{e}")).into_response(),
}
}
async fn api_nfs_shares_scan(
State(state): State<AppState>,
AxumPath(id): AxumPath<String>,
) -> Response {
match state.nfs_shares.rescan(&id).await {
Ok(n) => Json(json!({ "ok": true, "iso_count": n })).into_response(),
Err(e) => (StatusCode::BAD_REQUEST, format!("{e}")).into_response(),
}
}
// ─── Network info API ──────────────────────────────────────────────────────
async fn api_network(State(state): State<AppState>) -> Json<serde_json::Value> {
@@ -1922,11 +2039,21 @@ async fn api_metrics(State(state): State<AppState>) -> Response {
state
.metrics
.set_queue_counts(queue_entries.len() as u64, imaging as u64);
// v0.4.65: gauge tracks reachable external-storage shares. With
// NFS removed it now reflects SMB share reachability instead.
state
.metrics
.set_nfs_active(state.smb_shares.list().iter().filter(|m| m.reachable).count() as u64);
// v0.4.67: gauge tracks reachable external-storage shares
// across both protocols (SMB userspace + NFSv3 in-process).
let smb_ok = state
.smb_shares
.list()
.iter()
.filter(|m| m.reachable)
.count();
let nfs_ok = state
.nfs_shares
.list()
.iter()
.filter(|m| m.reachable)
.count();
state.metrics.set_nfs_active((smb_ok + nfs_ok) as u64);
let now = time::OffsetDateTime::now_utc();
let uptime = (now - state.started_at).whole_seconds().max(0) as u64;
+7 -1
View File
@@ -4,7 +4,7 @@ use openpxe_core::{
AdminStore, BootLog, BrandingStore, ClientRegistry, DeploymentQueue, HostBindings, LogBus,
Metrics, SettingsStore, SsoStore,
};
use openpxe_iso_store::{IsoStore, SmbManager, SmbShareManager};
use openpxe_iso_store::{IsoStore, NfsShareManager, SmbManager, SmbShareManager};
use std::sync::Arc;
use time::OffsetDateTime;
@@ -52,6 +52,12 @@ pub struct AppState {
/// `smbclient` does the SMB protocol over a plain TCP socket in
/// userspace — works in any container, no special caps required.
pub smb_shares: SmbShareManager,
/// v0.4.67: NFSv3 share manager — pure-Rust userspace consumer
/// via the `nfs3_client` crate. Ships alongside the SMB manager
/// so operators pick whichever protocol their NAS prefers.
/// In-process (no subprocess); supports HTTP Range requests on
/// NFS-sourced ISOs because NFSv3 READ3 takes an explicit offset.
pub nfs_shares: NfsShareManager,
/// Browser chunked upload state. Multipart uploads still go straight
/// through `IsoStore`, but the UI uses sessions so large ISO transfers
/// can show deterministic progress and leave visible partial files.
+112 -9
View File
@@ -88,12 +88,14 @@ async fn dispatch(state: &AppState, argv: &[String]) -> Result<String, String> {
"isos" | "images" => Ok(isos_text(state)),
"clients" => Ok(clients_text(state)),
"queue" => queue_command(state, tail).await,
// v0.4.65: `nfs` is gone — replaced with userspace SMB share
// consumer. `smb` still controls the outbound Samba server
// for Windows install media; `share` lists/manages remote SMB
// shares OpenPXE pulls ISOs from.
// `smb` controls the outbound Samba server for Windows
// install media. `share` lists/manages remote SMB shares
// OpenPXE pulls ISOs from (v0.4.65). `nfs` is the parallel
// command for remote NFSv3 shares (v0.4.67, in-process via
// nfs3_client — not the v0.4.64 kernel-mount path).
"share" | "smb-share" => smb_share_command(state, tail).await,
"smb" => smb_command(state, tail).await,
"nfs" => nfs_share_command(state, tail).await,
"log" => log_command(state, tail),
"whoami" => Ok("operator".to_string()),
"echo" => Ok(tail.join(" ")),
@@ -113,16 +115,20 @@ fn status_text(s: &AppState) -> String {
let smb = s.smb.as_ref().map(|m| m.snapshot());
let smb_shares = s.smb_shares.list();
let smb_reachable = smb_shares.iter().filter(|m| m.reachable).count();
// v0.4.67: NFSv3 sources too.
let nfs_shares = s.nfs_shares.list();
let nfs_reachable = nfs_shares.iter().filter(|m| m.reachable).count();
format!(
"OpenPXE {ver}\n\
base url: {base}\n\
interface: {nic}\n\
uptime: {up}\n\
isos: {n_isos} (local: {n_local}, smb: {n_smb})\n\
isos: {n_isos} (local: {n_local}, smb: {n_smb}, nfs: {n_nfs})\n\
clients: {n_clients}\n\
queue: {n_entries}\n\
smb server: {smb}\n\
smb shares: {n_total} configured ({n_active} reachable)\n",
smb shares: {n_smb_total} configured ({n_smb_active} reachable)\n\
nfs shares: {n_nfs_total} configured ({n_nfs_active} reachable)\n",
ver = env!("CARGO_PKG_VERSION"),
base = s.public_base_url,
nic = if s.nic_name.is_empty() {
@@ -140,11 +146,17 @@ fn status_text(s: &AppState) -> String {
.iter()
.filter(|i| matches!(i.source, openpxe_iso_store::IsoSource::Smb { .. }))
.count(),
n_nfs = isos
.iter()
.filter(|i| matches!(i.source, openpxe_iso_store::IsoSource::Nfs { .. }))
.count(),
n_clients = clients.len(),
n_entries = queue_entries.len(),
smb = smb.map_or_else(|| "(disabled)".into(), |s| format!("{s:?}")),
n_total = smb_shares.len(),
n_active = smb_reachable,
n_smb_total = smb_shares.len(),
n_smb_active = smb_reachable,
n_nfs_total = nfs_shares.len(),
n_nfs_active = nfs_reachable,
)
}
@@ -162,8 +174,9 @@ fn isos_text(s: &AppState) -> String {
for i in isos {
let src = match i.source {
openpxe_iso_store::IsoSource::Local => "local".to_string(),
// v0.4.65: SMB userspace consumer replaced kernel-mount NFS.
openpxe_iso_store::IsoSource::Smb { share_id, .. } => format!("smb:{share_id}"),
// v0.4.67: NFSv3 via in-process nfs3_client.
openpxe_iso_store::IsoSource::Nfs { share_id, .. } => format!("nfs:{share_id}"),
};
let _ = writeln!(
out,
@@ -378,6 +391,91 @@ async fn smb_share_command(s: &AppState, args: &[String]) -> Result<String, Stri
}
}
// ── nfs (v0.4.67: in-process NFSv3 via nfs3_client) ────────────────────
async fn nfs_share_command(s: &AppState, args: &[String]) -> Result<String, String> {
match args.first().map(String::as_str) {
None | Some("list") => {
let shares = s.nfs_shares.list();
if shares.is_empty() {
return Ok("(no NFS shares configured)".into());
}
let mut out = String::new();
let _ = writeln!(
out,
"{:<24} {:<7} {:<6} TARGET",
"ID", "STATUS", "ISOS"
);
for m in shares {
let status = if m.reachable { "ok" } else { "down" };
let _ = writeln!(
out,
"{:<24} {:<7} {:<6} {}:{}",
truncate(&m.id, 24),
status,
m.iso_count,
m.server,
m.export,
);
if let Some(e) = m.last_error {
let _ = writeln!(out, " error: {e}");
}
if let Some(h) = m.last_hint {
let _ = writeln!(out, " hint: {h}");
}
}
Ok(out)
}
Some("add") => {
// nfs add <server>:<export> [port]
let target = args
.get(1)
.ok_or_else(|| "usage: nfs add <server>:<export> [port]".to_string())?;
let (server, export) = target
.split_once(':')
.ok_or_else(|| "target must be 'server:/export'".to_string())?;
let port = args.get(2).and_then(|s| s.parse::<u16>().ok());
let req = openpxe_iso_store::NfsAddRequest {
server: server.to_string(),
export: export.to_string(),
port,
};
match s.nfs_shares.add(req).await {
Ok(m) => Ok(format!("added {} ({} isos)", m.id, m.iso_count)),
Err(e) => {
let mut out = format!("add failed: {}", e.error);
if let Some(h) = e.hint {
out.push_str("\nhint: ");
out.push_str(&h);
}
Err(out)
}
}
}
Some("remove") => {
let id = args
.get(1)
.ok_or_else(|| "usage: nfs remove <id>".to_string())?;
match s.nfs_shares.remove(id).await {
Ok(()) => Ok(format!("removed {id}")),
Err(e) => Err(format!("remove failed: {e}")),
}
}
Some("scan") => {
let id = args
.get(1)
.ok_or_else(|| "usage: nfs scan <id>".to_string())?;
match s.nfs_shares.rescan(id).await {
Ok(n) => Ok(format!("re-scanned {id}: {n} isos")),
Err(e) => Err(format!("scan failed: {e}")),
}
}
Some(other) => Err(format!(
"unknown nfs subcommand: {other}\ntry: nfs [list|add|remove|scan]"
)),
}
}
// ── smb ────────────────────────────────────────────────────────────────
#[allow(clippy::unused_async)]
@@ -519,6 +617,11 @@ OpenPXE terminal — available commands:
share remove <id> forget an SMB share
share scan <id> re-list a share for new ISOs
nfs list list configured NFSv3 shares
nfs add <srv>:<export> [port] add an NFSv3 share
nfs remove <id> forget an NFS share
nfs scan <id> re-list an NFS share for new ISOs
smb status outbound Samba state (Windows install media)
smb start | stop | reload control the outbound smbd
+51 -1
View File
@@ -14,7 +14,7 @@ use axum::body::Body;
use axum::http::{header, Request, StatusCode};
use openpxe_core::{ClientRegistry, DeploymentQueue, HostBindings, LogBus, Metrics, SettingsStore};
use openpxe_http_api::{build_router, AppState};
use openpxe_iso_store::{IsoStore, SmbShareManager};
use openpxe_iso_store::{IsoStore, NfsShareManager, SmbShareManager};
use tempfile::tempdir;
use tower::ServiceExt;
@@ -95,6 +95,7 @@ async fn build_state() -> (AppState, tempfile::TempDir) {
let queue = DeploymentQueue::new();
let settings = SettingsStore::load_or_default(dir.path());
let smb_shares = SmbShareManager::new(dir.path(), iso_store.clone());
let nfs_shares = NfsShareManager::new(dir.path(), iso_store.clone());
let log_bus = LogBus::new(64);
let hosts = HostBindings::load_or_default(dir.path());
let boot_log = openpxe_core::BootLog::load_or_default(dir.path());
@@ -117,6 +118,7 @@ async fn build_state() -> (AppState, tempfile::TempDir) {
metrics,
smb: None,
smb_shares,
nfs_shares,
uploads: openpxe_http_api::uploads::UploadSessions::default(),
log_bus,
started_at: time::OffsetDateTime::now_utc(),
@@ -531,6 +533,54 @@ async fn smb_shares_list_starts_empty() {
assert_eq!(v["shares"].as_array().unwrap().len(), 0);
}
// v0.4.67: NFSv3 share manager (parallel to SMB).
#[tokio::test]
async fn nfs_shares_list_starts_empty() {
let (state, _dir) = build_state().await;
let app = build_router(state);
let (s, b) = get(&app, "/api/nfs-shares").await;
assert_eq!(s, StatusCode::OK);
let v: serde_json::Value = serde_json::from_slice(&b).unwrap();
assert_eq!(v["shares"].as_array().unwrap().len(), 0);
}
#[tokio::test]
async fn nfs_shares_add_rejects_missing_server() {
let (state, _dir) = build_state().await;
let app = build_router(state);
let (s, b) = post_json(
&app,
"/api/nfs-shares",
r#"{"server":"","export":"/srv/isos"}"#,
)
.await;
assert_eq!(s, StatusCode::BAD_REQUEST);
let msg = String::from_utf8_lossy(&b);
assert!(
msg.to_lowercase().contains("server"),
"expected server hint, got: {msg}"
);
}
#[tokio::test]
async fn nfs_shares_add_rejects_export_without_leading_slash() {
let (state, _dir) = build_state().await;
let app = build_router(state);
let (s, b) = post_json(
&app,
"/api/nfs-shares",
r#"{"server":"10.0.0.5","export":"srv/isos"}"#,
)
.await;
assert_eq!(s, StatusCode::BAD_REQUEST);
let msg = String::from_utf8_lossy(&b);
assert!(
msg.to_lowercase().contains("export"),
"expected export-path hint, got: {msg}"
);
}
#[tokio::test]
async fn terminal_help_and_status_round_trip() {
let (state, _dir) = build_state().await;
+8
View File
@@ -35,5 +35,13 @@ libc = "0.2"
# encoder. Keeps the build slim — no JPEG2000, TIFF, BMP, etc.
image = { version = "0.25", default-features = false, features = ["png", "jpeg", "webp", "gif"] }
# v0.4.67: pure-Rust NFSv3 client for reading remote ISOs without a
# kernel mount. See crates/iso-store/src/nfs_share.rs for usage.
nfs3_client = { workspace = true }
nfs3_types = { workspace = true }
# Needed for the Stream trait that wraps the mpsc receiver feeding
# NFS read-loop bytes into axum's Body::from_stream.
futures = { workspace = true }
[dev-dependencies]
tempfile = "3.12"
+6
View File
@@ -18,6 +18,7 @@
pub mod entry;
pub mod introspect;
pub mod nfs_share;
pub mod pxe_logo;
pub mod smb;
pub mod smb_share;
@@ -33,6 +34,11 @@ pub use introspect::{DistroFamily, IntrospectionReport};
// it.
pub use smb::{extract_windows_iso, SmbManager, SmbState};
pub use smb_share::{SmbAddRequest, SmbShare, SmbShareError, SmbShareManager, SmbStream};
// v0.4.67: NFS is back — this time as an in-process userspace NFSv3
// client (the `nfs3_client` crate) rather than a kernel mount. Same
// "works in any container" property as SMB, plus support for HTTP
// Range requests because NFSv3 READ3 takes an explicit offset.
pub use nfs_share::{NfsAddRequest, NfsShare, NfsShareError, NfsShareManager, NfsStream};
pub use store::{
generate_boot_entries_for, slugify_str, IsoCategory, IsoMeta, IsoSource, IsoStore,
UploadHandle,
+917
View File
@@ -0,0 +1,917 @@
//! NFSv3 share consumer — in-process userspace client.
//!
//! v0.4.67 brings NFS back, this time *the right way*: a pure-Rust
//! NFSv3 client (`nfs3_client` from the xetdata family of crates)
//! that speaks the NFS protocol entirely in userspace over TCP. No
//! `mount.nfs`, no kernel modules, no `CAP_SYS_ADMIN`. Works in every
//! container the SMB path (v0.4.65) works in, including Unraid and
//! restricted-SCC OpenShift.
//!
//! ## How it differs from the v0.4.64 NFS path
//!
//! v0.4.64 shelled out to `mount(8)` and asked the kernel to attach
//! the remote share to the local filesystem. That required nfs client
//! modules on the **host** kernel. Containers that ran on hosts
//! without those modules — Unraid being the dominant case — failed
//! with `mount.nfs: failed to apply fstab options` no matter what
//! capabilities they were granted.
//!
//! v0.4.67 doesn't touch the kernel. The client opens a TCP socket to
//! the NFS server, performs the MOUNT3 RPC to get a root file handle,
//! and uses NFSv3 READDIR3 / LOOKUP3 / READ3 / GETATTR3 to walk the
//! share and stream files. The kernel sees plain TCP traffic, nothing
//! else.
//!
//! ## How it differs from the v0.4.65 SMB path
//!
//! SMB ships in v0.4.65 via the `smbclient` CLI subprocess. NFS ships
//! here via a Rust crate — no subprocess, no PATH lookup, no parsing
//! of human-formatted output. That also means:
//!
//! - **Range requests work** for NFS-sourced ISOs. NFSv3 READ3
//! takes an explicit offset, so the HTTP handler can seek into
//! the middle of a 5 GB ISO without downloading what comes
//! before. SMB-sourced ISOs still return 416 for ranges because
//! `smbclient -c 'get file -'` is a sequential stream.
//! - **Introspection could work** (we could LOOKUP the ISO9660 PVD
//! at offset 0x8000 and parse the volume label). For v0.4.67 we
//! don't yet — same as SMB, NFS-sourced ISOs register with an
//! `Unknown` family and fall back to generic sanboot. A follow-up
//! can add bounded read-based detection.
//!
//! ## How it's the same as SMB
//!
//! The public surface is deliberately parallel: `NfsShareManager`
//! mirrors `SmbShareManager`, `NfsShare` mirrors `SmbShare`, the API
//! responses use the same `{error, stderr, hint}` shape so the UI's
//! storage tab renders both protocols through one code path. The
//! state file (`<work_dir>/nfs_shares.json`) sits next to
//! `smb_shares.json`.
//!
//! ## No auth
//!
//! NFSv3 uses AUTH_SYS by default — the client tells the server "I'm
//! uid X, gid Y" and the server decides whether to trust that. Most
//! NAS appliances exposing ISO libraries via NFS gate access by
//! **client IP address** rather than uid, so there's nothing for the
//! UI to ask for. (If a future server needs Kerberos or non-default
//! uid mapping we can add those, but for ISO read access nobody does.)
use crate::introspect::{DistroFamily, IntrospectionReport};
use crate::store::{generate_boot_entries_for, slugify_str, IsoSource, IsoStore};
use bytes::Bytes;
use nfs3_client::tokio::TokioConnector;
use nfs3_client::Nfs3ConnectionBuilder;
use nfs3_types::nfs3::{
self as nfs3, diropargs3, entry3, filename3, nfs_fh3, GETATTR3args, LOOKUP3args,
Nfs3Result, READ3args, READDIR3args,
};
use nfs3_types::xdr_codec::Opaque;
use openpxe_core::{Error, Result};
use parking_lot::Mutex;
use serde::{Deserialize, Serialize};
use std::borrow::Cow;
use std::collections::HashMap;
use std::path::{Path, PathBuf};
use std::sync::Arc;
use std::time::Duration;
use time::OffsetDateTime;
/// Default TCP port for the NFSv3 protocol. The classic split (111
/// portmapper + dynamic mountd) is supported by the underlying crate;
/// for direct connection-mode NAS appliances 2049 is what answers.
const DEFAULT_NFS_PORT: u16 = 2049;
/// How long we wait for the initial MOUNT3 + portmap handshake before
/// giving up. Mirrors the SMB pre-flight probe shape so the UI's
/// timeout banner reads consistently across protocols.
const CONNECT_TIMEOUT: Duration = Duration::from_secs(8);
/// NFSv3 READ3 chunk size. The protocol's hard cap is 1 MiB per
/// reply; 64 KiB is a polite default that nearly every server hands
/// back without fragmentation and keeps in-flight memory bounded
/// during a streaming response.
const READ_CHUNK_BYTES: u32 = 64 * 1024;
/// Bound on the in-flight queue between the read-loop task and the
/// HTTP body stream. 16 * 64 KiB ≈ 1 MiB max buffer per stream —
/// enough to keep the network pipe full without letting a slow
/// client park gigabytes of decoded ISO in RAM.
const STREAM_BUFFER_DEPTH: usize = 16;
/// One configured NFS share. The id is derived from server+export so
/// re-adding the same coordinates is idempotent.
#[derive(Debug, Clone, Serialize, Deserialize)]
pub struct NfsShare {
pub id: String,
pub server: String,
/// Export path on the server (e.g. "/srv/isos"). Must start with
/// "/" to match the NFS server's view; we validate on add.
pub export: String,
/// TCP port — 2049 unless the operator overrode it.
#[serde(default = "default_port")]
pub port: u16,
/// Most recent error talking to the share, or `None` on success.
pub last_error: Option<String>,
/// Operator-friendly translation of `last_error`. None when we
/// don't have a friendlier rendition.
pub last_hint: Option<String>,
#[serde(with = "time::serde::rfc3339::option")]
pub last_scan: Option<OffsetDateTime>,
pub iso_count: u32,
/// True after a successful scan, false on failure. Drives the
/// UI dot.
pub reachable: bool,
}
/// Submission from the UI / API.
#[derive(Debug, Clone, Deserialize)]
pub struct NfsAddRequest {
pub server: String,
pub export: String,
#[serde(default)]
pub port: Option<u16>,
}
fn default_port() -> u16 {
DEFAULT_NFS_PORT
}
/// Structured error surfaced to the API and rendered in the UI. Same
/// shape as `SmbShareError` so the storage tab uses one rendering
/// path for both protocols.
#[derive(Debug, Clone, Serialize)]
pub struct NfsShareError {
pub error: String,
pub stderr: String,
pub hint: Option<String>,
}
impl NfsShareError {
fn from_raw(error: impl Into<String>, stderr: impl Into<String>) -> Self {
let stderr = stderr.into();
let error = error.into();
let hint = hint_for(&stderr).or_else(|| hint_for(&error));
Self {
error,
stderr,
hint,
}
}
}
#[derive(Debug, Default)]
struct Inner {
shares: HashMap<String, NfsShare>,
}
/// Manages NFS shares. Cheap to clone — internal state is
/// `Arc<Mutex<...>>`.
#[derive(Debug, Clone)]
pub struct NfsShareManager {
state_path: Arc<PathBuf>,
inner: Arc<Mutex<Inner>>,
iso_store: IsoStore,
/// Serializes scan operations on the same manager. Each scan
/// opens its own NFS connection so concurrency isn't a hard
/// requirement, but serializing keeps log output predictable.
op_lock: Arc<tokio::sync::Mutex<()>>,
}
impl NfsShareManager {
/// Construct a manager. State persists to
/// `<work_dir>/nfs_shares.json`.
#[must_use]
pub fn new(work_dir: &Path, iso_store: IsoStore) -> Self {
let state_path = work_dir.join("nfs_shares.json");
Self {
state_path: Arc::new(state_path),
inner: Arc::new(Mutex::new(Inner::default())),
iso_store,
op_lock: Arc::new(tokio::sync::Mutex::new(())),
}
}
/// Load persisted state and re-scan every share. Per-share
/// failures are logged but never propagated — startup must not
/// block on a single offline server.
pub async fn load_and_rescan(&self) -> Result<()> {
let shares = match tokio::fs::read_to_string(self.state_path.as_path()).await {
Ok(text) => serde_json::from_str::<Vec<NfsShare>>(&text).unwrap_or_default(),
Err(_) => Vec::new(),
};
for mut s in shares {
s.last_error = None;
s.last_hint = None;
s.reachable = false;
self.inner.lock().shares.insert(s.id.clone(), s.clone());
if let Err(e) = self.rescan_inner(&s.id).await {
tracing::warn!(
target: "openpxe::nfs",
id = %s.id, server = %s.server, export = %s.export,
"rescan on startup failed: {e}"
);
}
}
Ok(())
}
/// Register an NFS share. Validates, probes connectivity by
/// performing a real MOUNT3 + READDIR3, and registers the
/// resulting ISOs with the store.
pub async fn add(
&self,
req: NfsAddRequest,
) -> std::result::Result<NfsShare, NfsShareError> {
let server = normalize_server(&req.server);
let export = req.export.trim().to_string();
if server.is_empty() {
return Err(NfsShareError::from_raw("server is required", ""));
}
if !export.starts_with('/') {
return Err(NfsShareError::from_raw(
"export path must start with '/' (e.g. /srv/isos)",
"",
));
}
if server.contains('\0') || export.contains('\0') {
return Err(NfsShareError::from_raw("NUL bytes are not allowed", ""));
}
let port = req.port.filter(|p| *p != 0).unwrap_or(DEFAULT_NFS_PORT);
let id = share_id(&server, &export);
let spec = NfsShare {
id: id.clone(),
server,
export,
port,
last_error: None,
last_hint: None,
last_scan: None,
iso_count: 0,
reachable: false,
};
self.inner.lock().shares.insert(id.clone(), spec);
self.persist_locked();
if let Err(e) = self.rescan_inner(&id).await {
let m = self.get(&id);
return Err(NfsShareError {
error: m.as_ref().and_then(|m| m.last_error.clone())
.unwrap_or_else(|| e.to_string()),
stderr: String::new(),
hint: m.and_then(|m| m.last_hint),
});
}
Ok(self.get(&id).expect("just inserted"))
}
/// Remove a share. Drops every ISO sourced from it. Idempotent.
/// Async for symmetry with [`SmbShareManager::remove`] — the
/// SMB version is async because it tears down a credentials
/// file; NFS has nothing to clean up but we keep the signature
/// uniform so the call sites in app.rs / terminal.rs match.
#[allow(clippy::unused_async)]
pub async fn remove(&self, id: &str) -> Result<()> {
let removed = self.inner.lock().shares.remove(id).is_some();
if removed {
self.iso_store.drop_external_source(id);
self.persist_locked();
}
Ok(())
}
/// Re-walk a share for new / removed ISOs.
pub async fn rescan(&self, id: &str) -> Result<u32> {
self.rescan_inner(id).await
}
#[must_use]
pub fn list(&self) -> Vec<NfsShare> {
let g = self.inner.lock();
let mut v: Vec<_> = g.shares.values().cloned().collect();
v.sort_by(|a, b| a.id.cmp(&b.id));
v
}
#[must_use]
pub fn get(&self, id: &str) -> Option<NfsShare> {
self.inner.lock().shares.get(id).cloned()
}
/// Open a byte stream reading `filename` out of share `share_id`,
/// starting at `start_offset` and reading at most `max_len`
/// bytes. The returned stream yields `Bytes` chunks (≤
/// `READ_CHUNK_BYTES`) and ends at EOF, after `max_len` bytes, or
/// on the first transport error.
///
/// Used by the HTTP ISO download handler. Supports HTTP Range
/// requests because NFSv3 READ3 takes an explicit offset — this
/// is the protocol-level advantage NFS has over the SMB
/// userspace path.
///
/// Kept `async` for symmetry with [`SmbShareManager::stream_iso`]
/// even though the body doesn't await today — a future
/// refinement (e.g. throttling, connection pooling) will need to
/// await without changing call sites.
#[allow(clippy::unused_async)]
pub async fn stream_iso(
&self,
share_id: &str,
filename: &str,
start_offset: u64,
max_len: Option<u64>,
) -> Result<NfsStream> {
let share = self
.get(share_id)
.ok_or_else(|| Error::Invalid(format!("no such NFS share '{share_id}'")))?;
// Defensive: NFSv3 LOOKUP3 takes a single name relative to
// the export root, not a path. We don't support nested
// directories in v0.4.67 — ISOs live at the top of the share.
if filename.contains('/') || filename.contains('\\') || filename.contains("..") {
return Err(Error::Invalid(format!("invalid filename '{filename}'")));
}
let (tx, rx) =
tokio::sync::mpsc::channel::<std::io::Result<Bytes>>(STREAM_BUFFER_DEPTH);
let server = share.server.clone();
let export = share.export.clone();
let port = share.port;
let fname = filename.to_string();
// Spawn a task that owns the NFS connection. Owning the
// connection inside the spawn means we don't have to worry
// about borrowing across awaits or sharing the connection
// between scan and stream — each stream gets its own.
let task = tokio::spawn(async move {
let result = stream_loop(
&server,
&export,
port,
&fname,
start_offset,
max_len,
tx.clone(),
)
.await;
if let Err(e) = result {
// Best-effort signal of the error to the consumer.
// If the receiver has already dropped we just exit.
let _ = tx
.send(Err(std::io::Error::other(e.to_string())))
.await;
}
});
Ok(NfsStream {
rx,
_task: task,
})
}
// ── internals ─────────────────────────────────────────────────────
async fn rescan_inner(&self, id: &str) -> Result<u32> {
let _g = self.op_lock.lock().await;
let share = self
.get(id)
.ok_or_else(|| Error::Invalid(format!("no such share '{id}'")))?;
let now = OffsetDateTime::now_utc();
// Drop prior entries so a deleted file disappears from the
// store on the next scan.
self.iso_store.drop_external_source(id);
let listing = match list_isos(&share.server, &share.export, share.port).await {
Ok(l) => l,
Err(err) => {
let stderr = err.to_string();
let hint = hint_for(&stderr);
self.update_status(id, 0, false, Some(stderr.clone()), hint, now);
return Err(Error::Invalid(stderr));
}
};
let mut count = 0u32;
for entry in listing {
let iso_id = format!("nfs-{}-{}", share.id, slugify_str(&entry.filename));
// Same approach as SMB: no real introspection over the
// network in v0.4.67. The boot-entry generator falls back
// to filename-based sanboot detection.
let report = IntrospectionReport {
family: DistroFamily::Unknown,
volume_label: None,
kernel_path: None,
initrd_paths: Vec::new(),
has_boot_wim: false,
};
let boot_entries = generate_boot_entries_for(&iso_id, &entry.filename, &report);
let source = IsoSource::Nfs {
share_id: share.id.clone(),
relative_path: entry.filename.clone(),
};
self.iso_store.register_external(
iso_id,
entry.filename,
entry.size,
report,
boot_entries,
source,
);
count += 1;
}
self.update_status(id, count, true, None, None, now);
tracing::info!(
target: "openpxe::nfs",
id = %id, server = %share.server, export = %share.export,
iso_count = count,
"NFS share scanned"
);
Ok(count)
}
fn update_status(
&self,
id: &str,
iso_count: u32,
reachable: bool,
err: Option<String>,
hint: Option<String>,
ts: OffsetDateTime,
) {
if let Some(s) = self.inner.lock().shares.get_mut(id) {
s.iso_count = iso_count;
s.reachable = reachable;
s.last_error = err;
s.last_hint = hint;
s.last_scan = Some(ts);
}
self.persist_locked();
}
fn persist_locked(&self) {
let shares: Vec<NfsShare> = self.inner.lock().shares.values().cloned().collect();
let path = self.state_path.as_path();
let tmp = path.with_extension("json.tmp");
let body = match serde_json::to_vec_pretty(&shares) {
Ok(b) => b,
Err(e) => {
tracing::warn!(target: "openpxe::nfs", "serialize: {e}");
return;
}
};
if let Some(parent) = path.parent() {
let _ = std::fs::create_dir_all(parent);
}
if let Err(e) = std::fs::write(&tmp, body) {
tracing::warn!(target: "openpxe::nfs", "write tmp: {e}");
return;
}
if let Err(e) = std::fs::rename(&tmp, path) {
tracing::warn!(target: "openpxe::nfs", "rename: {e}");
}
}
}
/// HTTP body stream for an NFS-sourced ISO read.
///
/// Implements `Stream<Item = io::Result<Bytes>>` so axum can convert
/// it into a response body via `Body::from_stream`.
#[derive(Debug)]
pub struct NfsStream {
rx: tokio::sync::mpsc::Receiver<std::io::Result<Bytes>>,
/// Kept alive so the read-loop task isn't dropped while the HTTP
/// client is still consuming bytes. Dropping the stream cancels
/// the task, which is the right behavior on client disconnect.
_task: tokio::task::JoinHandle<()>,
}
impl futures::Stream for NfsStream {
type Item = std::io::Result<Bytes>;
fn poll_next(
mut self: std::pin::Pin<&mut Self>,
cx: &mut std::task::Context<'_>,
) -> std::task::Poll<Option<Self::Item>> {
self.rx.poll_recv(cx)
}
}
#[derive(Debug, Clone)]
struct NfsListEntry {
filename: String,
size: u64,
}
/// Connect, READDIR the export root, look up each `*.iso` to get its
/// size + file handle. Returns a flat list. Errors are returned with
/// a human-readable message; the caller decides how to surface them.
async fn list_isos(
server: &str,
export: &str,
port: u16,
) -> std::result::Result<Vec<NfsListEntry>, NfsClientError> {
let mut conn =
tokio::time::timeout(CONNECT_TIMEOUT, build_connection(server, export, port))
.await
.map_err(|_| NfsClientError::Timeout(server.to_string(), port))??;
let root = conn.root_nfs_fh3();
let mut entries = Vec::new();
let mut cookie: u64 = 0;
let mut cookieverf = nfs3::cookieverf3::default();
loop {
let res = conn
.readdir(&READDIR3args {
dir: root.clone(),
cookie,
cookieverf,
count: 32 * 1024,
})
.await
.map_err(NfsClientError::Rpc)?;
let ok = match res {
Nfs3Result::Ok(ok) => ok,
Nfs3Result::Err((status, _)) => {
return Err(NfsClientError::Nfsstat(status_label(status)));
}
};
cookieverf = ok.cookieverf;
let eof = ok.reply.eof;
let dir_entries: Vec<entry3<'_>> = ok.reply.entries.0;
let next_cookie = dir_entries.last().map(|e| e.cookie);
for entry in dir_entries {
// entry.name is `filename3<'a>(Opaque<'a>)` — XDR opaque
// bytes. Decode to UTF-8 best-effort.
let name_bytes = entry.name.0.as_ref();
let Ok(name) = std::str::from_utf8(name_bytes) else {
continue;
};
if name == "." || name == ".." {
continue;
}
if !name.to_ascii_lowercase().ends_with(".iso") {
continue;
}
// Look up the file to get its size + verify it's a
// regular file (not a symlink / directory matching the
// .iso pattern).
let lookup = conn
.lookup(&LOOKUP3args {
what: diropargs3 {
dir: root.clone(),
name: filename3(Opaque::borrowed(name_bytes)),
},
})
.await
.map_err(NfsClientError::Rpc)?;
let lookup_ok = match lookup {
Nfs3Result::Ok(o) => o,
Nfs3Result::Err(_) => continue,
};
let getattr = conn
.getattr(&GETATTR3args {
object: lookup_ok.object,
})
.await
.map_err(NfsClientError::Rpc)?;
let attrs = match getattr {
Nfs3Result::Ok(o) => o.obj_attributes,
Nfs3Result::Err(_) => continue,
};
// ftype3::NF3REG == 1 (regular file). Skip everything
// else — directories, symlinks, devices.
if attrs.type_ as u32 != nfs3::ftype3::NF3REG as u32 {
continue;
}
entries.push(NfsListEntry {
filename: name.to_string(),
size: attrs.size,
});
}
if eof {
break;
}
match next_cookie {
Some(c) if c != 0 => cookie = c,
_ => break,
}
}
let _ = conn.unmount().await;
Ok(entries)
}
async fn stream_loop(
server: &str,
export: &str,
port: u16,
filename: &str,
start_offset: u64,
max_len: Option<u64>,
tx: tokio::sync::mpsc::Sender<std::io::Result<Bytes>>,
) -> std::result::Result<(), NfsClientError> {
let mut conn =
tokio::time::timeout(CONNECT_TIMEOUT, build_connection(server, export, port))
.await
.map_err(|_| NfsClientError::Timeout(server.to_string(), port))??;
let root = conn.root_nfs_fh3();
// Look up the file to get its handle.
let name_bytes = filename.as_bytes();
let lookup = conn
.lookup(&LOOKUP3args {
what: diropargs3 {
dir: root,
name: filename3(Opaque::borrowed(name_bytes)),
},
})
.await
.map_err(NfsClientError::Rpc)?;
let lookup_ok = match lookup {
Nfs3Result::Ok(o) => o,
Nfs3Result::Err((status, _)) => {
return Err(NfsClientError::Nfsstat(status_label(status)));
}
};
let file_handle: nfs_fh3 = lookup_ok.object;
let mut offset = start_offset;
let mut remaining = max_len;
loop {
if let Some(r) = remaining {
if r == 0 {
break;
}
}
// Cap the chunk at READ_CHUNK_BYTES and at the remaining budget.
let chunk = match remaining {
Some(r) if r < u64::from(READ_CHUNK_BYTES) => r as u32,
_ => READ_CHUNK_BYTES,
};
let res = conn
.read(&READ3args {
file: file_handle.clone(),
offset,
count: chunk,
})
.await
.map_err(NfsClientError::Rpc)?;
let ok = match res {
Nfs3Result::Ok(o) => o,
Nfs3Result::Err((status, _)) => {
return Err(NfsClientError::Nfsstat(status_label(status)));
}
};
let bytes = Bytes::copy_from_slice(ok.data.as_ref());
let bytes_len = bytes.len() as u64;
if tx.send(Ok(bytes)).await.is_err() {
// HTTP client dropped — abort gracefully.
break;
}
offset += bytes_len;
if let Some(r) = remaining.as_mut() {
*r = r.saturating_sub(bytes_len);
}
if ok.eof {
break;
}
// Defensive: a server that returns 0 bytes without EOF
// would have us busy-looping. Bail out instead.
if bytes_len == 0 {
break;
}
}
let _ = conn.unmount().await;
Ok(())
}
/// Hand the connection builder the user's settings. `mount_path` is
/// the server-side export (e.g. "/srv/isos"). We disable
/// `connect_from_privileged_port` because the openpxe process runs
/// as uid 10001 and can't bind sub-1024 source ports — and most
/// modern NFS servers no longer require them anyway. If a server
/// does demand it the operator's hint will guide them to the
/// `insecure` export option.
async fn build_connection(
server: &str,
export: &str,
port: u16,
) -> std::result::Result<
nfs3_client::Nfs3Connection<nfs3_client::tokio::TokioIo<tokio::net::TcpStream>>,
NfsClientError,
> {
Nfs3ConnectionBuilder::new(TokioConnector, server, export)
.connect_from_privileged_port(false)
.nfs3_port(port)
.mount()
.await
.map_err(|e| NfsClientError::Connect(e.to_string()))
}
#[derive(Debug)]
enum NfsClientError {
Connect(String),
Timeout(String, u16),
Rpc(nfs3_client::RpcError),
Nfsstat(String),
}
impl std::fmt::Display for NfsClientError {
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
match self {
Self::Connect(msg) => write!(f, "connect failed: {msg}"),
Self::Timeout(host, port) => write!(
f,
"connect timed out after {}s talking to {host}:{port}",
CONNECT_TIMEOUT.as_secs()
),
Self::Rpc(e) => write!(f, "RPC failure: {e}"),
Self::Nfsstat(s) => write!(f, "NFS server returned {s}"),
}
}
}
impl std::error::Error for NfsClientError {}
/// Best-effort label for an `nfsstat3` so the UI shows a readable
/// name like `NFS3ERR_ACCES` instead of a magic number.
fn status_label(status: nfs3::nfsstat3) -> String {
use nfs3::nfsstat3 as S;
let name = match status {
S::NFS3_OK => "NFS3_OK",
S::NFS3ERR_PERM => "NFS3ERR_PERM",
S::NFS3ERR_NOENT => "NFS3ERR_NOENT",
S::NFS3ERR_IO => "NFS3ERR_IO",
S::NFS3ERR_NXIO => "NFS3ERR_NXIO",
S::NFS3ERR_ACCES => "NFS3ERR_ACCES",
S::NFS3ERR_EXIST => "NFS3ERR_EXIST",
S::NFS3ERR_XDEV => "NFS3ERR_XDEV",
S::NFS3ERR_NODEV => "NFS3ERR_NODEV",
S::NFS3ERR_NOTDIR => "NFS3ERR_NOTDIR",
S::NFS3ERR_ISDIR => "NFS3ERR_ISDIR",
S::NFS3ERR_INVAL => "NFS3ERR_INVAL",
S::NFS3ERR_FBIG => "NFS3ERR_FBIG",
S::NFS3ERR_NOSPC => "NFS3ERR_NOSPC",
S::NFS3ERR_ROFS => "NFS3ERR_ROFS",
S::NFS3ERR_MLINK => "NFS3ERR_MLINK",
S::NFS3ERR_NAMETOOLONG => "NFS3ERR_NAMETOOLONG",
S::NFS3ERR_NOTEMPTY => "NFS3ERR_NOTEMPTY",
S::NFS3ERR_DQUOT => "NFS3ERR_DQUOT",
S::NFS3ERR_STALE => "NFS3ERR_STALE",
S::NFS3ERR_REMOTE => "NFS3ERR_REMOTE",
S::NFS3ERR_BADHANDLE => "NFS3ERR_BADHANDLE",
S::NFS3ERR_NOT_SYNC => "NFS3ERR_NOT_SYNC",
S::NFS3ERR_BAD_COOKIE => "NFS3ERR_BAD_COOKIE",
S::NFS3ERR_NOTSUPP => "NFS3ERR_NOTSUPP",
S::NFS3ERR_TOOSMALL => "NFS3ERR_TOOSMALL",
S::NFS3ERR_SERVERFAULT => "NFS3ERR_SERVERFAULT",
S::NFS3ERR_BADTYPE => "NFS3ERR_BADTYPE",
S::NFS3ERR_JUKEBOX => "NFS3ERR_JUKEBOX",
};
name.to_string()
}
/// Translate well-known NFS error strings into actionable hints for
/// the UI. Mirrors the SMB hint table in spirit; the patterns are
/// different because NFS errors travel via NFS3ERR_* codes plus
/// transport-level messages from the Rust crate.
fn hint_for(text: &str) -> Option<String> {
let s = text.to_ascii_lowercase();
if s.contains("nfs3err_acces") || s.contains("permission denied") {
Some(
"the NFS server rejected this client. Most likely your export \
is restricted by client IP — add this OpenPXE host (or its \
subnet) to the export's allowed-clients list on the server."
.into(),
)
} else if s.contains("nfs3err_noent") || s.contains("nfs3err_notdir") {
Some(
"the export path doesn't exist on the server, or it isn't a \
directory. Double-check the path (e.g. /srv/isos vs /isos — \
UniFi UNAS Pro exposes shares under /var/nfs/shared/<name>)."
.into(),
)
} else if s.contains("nfs3err_stale") || s.contains("nfs3err_badhandle") {
Some(
"the server's view of the share changed under us. Re-scan; \
if that doesn't help, remove and re-add the share."
.into(),
)
} else if s.contains("connect timed out")
|| s.contains("timed out")
|| s.contains("connection timed out")
{
Some(
"no answer from the server within the connect timeout. \
Verify the IP, the port (default 2049), and any firewall \
between OpenPXE and the NAS."
.into(),
)
} else if s.contains("connection refused") {
Some(
"the NFS service isn't accepting connections on this port. \
Make sure nfsd is running on the server and (for v3) the \
portmapper on port 111 is reachable."
.into(),
)
} else if s.contains("no route to host") || s.contains("network is unreachable") {
Some(
"the server isn't reachable on this network. Check the IP \
and routes."
.into(),
)
} else if s.contains("mount") && s.contains("denied") {
Some(
"MOUNT3 was denied. The classic cause is that the export's \
`rw=<host>` / `ro=<host>` list doesn't include this client. \
Some servers also require `insecure` in /etc/exports for \
non-privileged source ports — which is what this client \
uses (we run as uid 10001, no privileged-port capability)."
.into(),
)
} else {
None
}
}
fn share_id(server: &str, export: &str) -> String {
slugify_str(&format!("{server}{export}"))
}
/// Normalize a server input: trim, strip schemes, drop trailing
/// slashes. Matches the SMB normalizer so paste-from-anywhere works.
fn normalize_server(raw: &str) -> String {
let s = raw.trim();
let s = s
.strip_prefix("nfs://")
.or_else(|| s.strip_prefix("http://"))
.or_else(|| s.strip_prefix("https://"))
.unwrap_or(s);
s.trim_end_matches('/').to_string()
}
// Silence the unused-import warning on `Cow` — we use it implicitly
// via `Opaque::borrowed` constructions in `list_isos`.
#[allow(dead_code)]
fn _unused() -> Cow<'static, str> {
Cow::Borrowed("")
}
#[cfg(test)]
mod tests {
use super::*;
#[test]
fn share_id_is_stable_and_safe() {
let a = share_id("10.0.0.5", "/srv/isos");
let b = share_id("10.0.0.5", "/srv/isos");
assert_eq!(a, b);
assert!(!a.contains('/'));
assert!(!a.contains('.'));
}
#[test]
fn normalize_server_strips_schemes() {
assert_eq!(normalize_server(" 10.0.0.5 "), "10.0.0.5");
assert_eq!(normalize_server("nfs://nas.lan/"), "nas.lan");
assert_eq!(normalize_server("http://192.168.1.51"), "192.168.1.51");
assert_eq!(normalize_server("nas.lan"), "nas.lan");
}
#[test]
fn hint_for_acces_points_to_exports_table() {
let h = hint_for("NFS server returned NFS3ERR_ACCES").unwrap();
assert!(
h.to_lowercase().contains("export"),
"expected exports guidance, got: {h}"
);
}
#[test]
fn hint_for_noent_points_to_export_path() {
let h = hint_for("NFS server returned NFS3ERR_NOENT").unwrap();
assert!(
h.to_lowercase().contains("path") || h.to_lowercase().contains("directory"),
"expected export-path guidance, got: {h}"
);
}
#[test]
fn hint_for_mount_denied_calls_out_insecure_option() {
let h = hint_for("mount denied").unwrap();
assert!(h.to_lowercase().contains("insecure"));
}
#[test]
fn hint_for_unknown_is_none() {
assert!(hint_for("some entirely unrelated string").is_none());
}
#[test]
fn status_label_covers_common_codes() {
assert_eq!(status_label(nfs3::nfsstat3::NFS3_OK), "NFS3_OK");
assert_eq!(status_label(nfs3::nfsstat3::NFS3ERR_ACCES), "NFS3ERR_ACCES");
assert_eq!(status_label(nfs3::nfsstat3::NFS3ERR_NOENT), "NFS3ERR_NOENT");
}
}
+32 -19
View File
@@ -15,27 +15,33 @@ use tokio::io::AsyncWriteExt;
/// Where the bytes for an ISO actually live.
///
/// The default is `Local` — uploaded ISOs sit in `<iso_dir>/<id>.iso`.
/// `Smb` entries (v0.4.65) point at a file inside a remote SMB share
/// that the `SmbShareManager` knows how to stream via Samba's
/// userspace `smbclient` CLI. The HTTP handler resolves the share by
/// id at request time and pipes `smbclient -c 'get file -'` straight
/// into the response body — no kernel mount, no local cache.
/// `Local` — uploaded ISO, sits at `<iso_dir>/<id>.iso`.
/// `Smb` (v0.4.65) — remote SMB share, streamed via Samba's
/// userspace `smbclient` CLI subprocess. No kernel mount, no local
/// cache. Sequential whole-file streaming; HTTP Range requests
/// return 416.
/// `Nfs` (v0.4.67) — remote NFSv3 share, streamed via the pure-Rust
/// `nfs3_client` crate (in-process, no subprocess). Same "works in
/// any container" property as SMB, plus Range requests work because
/// NFSv3 READ3 takes an explicit offset.
#[derive(Debug, Clone, Default, Serialize, Deserialize)]
#[serde(tag = "kind", rename_all = "snake_case")]
pub enum IsoSource {
#[default]
Local,
/// v0.4.65: kernel-mount NFS is gone (it didn't work on Unraid
/// regardless of capabilities — the host kernel needs the nfs
/// client modules loaded). SMB via userspace `smbclient` works in
/// any container.
/// v0.4.65: SMB via userspace `smbclient` works in any container.
Smb {
share_id: String,
/// Filename at the share root. We don't support nested paths
/// in v0.4.65; ISOs live at the top of the share.
relative_path: String,
},
/// v0.4.67: NFSv3 via the in-process `nfs3_client` crate.
Nfs {
share_id: String,
/// Filename at the export root.
relative_path: String,
},
}
/// Where the ISO lands in the PXE menu hierarchy.
@@ -293,10 +299,11 @@ impl IsoStore {
None
}
}
// SMB sources have no local path — they're streamed via
// smbclient subprocess. Callers should check the source
// kind first and dispatch accordingly.
IsoSource::Smb { .. } => None,
// SMB and NFS sources have no local path — they're
// streamed in-process. Callers must inspect the source
// kind first and dispatch to the appropriate share
// manager.
IsoSource::Smb { .. } | IsoSource::Nfs { .. } => None,
}
}
@@ -348,13 +355,19 @@ impl IsoStore {
}
/// Drop every entry that belongs to `share_id`. Used by the SMB
/// share manager when an operator removes a share, or before
/// re-scanning to clean out stale entries.
/// and NFS share managers when an operator removes a share, or
/// before re-scanning to clean out stale entries. The same id
/// space serves both protocols — share ids are slugified from
/// `server+share` (SMB) or `server+export` (NFS) and the
/// protocol-specific prefix prevents collisions.
pub fn drop_external_source(&self, share_id: &str) {
let mut g = self.inner.write();
g.isos.retain(
|_, m| !matches!(&m.source, IsoSource::Smb { share_id: sid, .. } if sid == share_id),
);
g.isos.retain(|_, m| match &m.source {
IsoSource::Smb { share_id: sid, .. } | IsoSource::Nfs { share_id: sid, .. } => {
sid != share_id
}
IsoSource::Local => true,
});
}
/// Set or clear an ISO's boot password.
+14 -1
View File
@@ -9,7 +9,7 @@ use openpxe_core::{
};
use openpxe_dhcp_proxy::DhcpProxyServer;
use openpxe_http_api::{build_router, AppState};
use openpxe_iso_store::{IsoStore, SmbManager, SmbShareManager};
use openpxe_iso_store::{IsoStore, NfsShareManager, SmbManager, SmbShareManager};
use openpxe_tftp::TftpServer;
use std::net::{Ipv4Addr, SocketAddr};
use std::path::PathBuf;
@@ -133,6 +133,18 @@ async fn main() -> anyhow::Result<()> {
);
}
// v0.4.67: NFSv3 share manager — pure-Rust in-process consumer
// via the `nfs3_client` crate. Sits alongside the SMB manager;
// operators pick whichever protocol their NAS prefers, or use
// both. No subprocess, no kernel mount, works in any container.
let nfs_shares = NfsShareManager::new(&config.paths.work_dir, iso_store.clone());
if let Err(e) = nfs_shares.load_and_rescan().await {
tracing::warn!(
target: "openpxe::nfs",
"could not reload NFS shares on startup: {e}"
);
}
// Sniff network details for the Network tab. None of these are
// required for PXE to work — they're informational, surfaced in the
// UI so an operator doesn't have to drop to a shell to find their
@@ -158,6 +170,7 @@ async fn main() -> anyhow::Result<()> {
metrics: metrics.clone(),
smb: Some(smb.clone()),
smb_shares: smb_shares.clone(),
nfs_shares: nfs_shares.clone(),
uploads: openpxe_http_api::uploads::UploadSessions::default(),
log_bus: log_bus.clone(),
started_at: time::OffsetDateTime::now_utc(),
+127 -15
View File
@@ -167,8 +167,11 @@
el('div', {class: 'trend'},
isos.filter(i => i.introspection.family === 'windows_pe').length + ' Windows · ' +
isos.filter(i => i.introspection.family !== 'windows_pe').length + ' Linux · ' +
(status.smb_share_reachable || 0) + ' SMB share' +
((status.smb_share_reachable || 0) === 1 ? '' : 's')),
// v0.4.67: count both protocols. Label generically since
// operators may be using one, the other, or both.
((status.smb_share_reachable || 0) + (status.nfs_share_reachable || 0)) +
' remote share' +
(((status.smb_share_reachable || 0) + (status.nfs_share_reachable || 0)) === 1 ? '' : 's')),
])),
el('div', {class: 'card'}, el('div', {class: 'stat'}, [
el('div', {class: 'label'}, 'Uptime'),
@@ -345,16 +348,20 @@
storage: async () => {
// v0.4.65: kernel-mount NFS replaced with userspace SMB via
// smbclient — works in any container regardless of host kernel
// modules or capabilities. The /api/nfs endpoint is gone;
// /api/smb-shares is the replacement.
const [isos, settings, smbRes, disk] = await Promise.all([
// modules or capabilities.
// v0.4.67: NFSv3 added back as an in-process Rust client
// (nfs3_client crate). Both protocols available side-by-side;
// operators pick whichever their NAS prefers.
const [isos, settings, smbRes, nfsRes, disk] = await Promise.all([
getJSON('/api/isos'), getJSON('/api/settings'),
getJSON('/api/smb-shares'),
getJSON('/api/nfs-shares'),
getJSON('/api/storage/disk').catch(() => ({
total_bytes: 0, available_bytes: 0, used_bytes: 0, path: '?',
})),
]);
const shares = smbRes.shares || [];
const nfsShares = nfsRes.shares || [];
// ── Upload card ──
const drop = el('div', {class:'drop', id:'drop'}, [
@@ -464,10 +471,14 @@
const rowsAndEditors = [];
isos.forEach(i => {
const b = bootability(i, settings);
// v0.4.65: SMB userspace consumer replaced NFS. The badge
// colours stay the same so the table looks unchanged for
// existing operators.
// v0.4.65: SMB userspace consumer (smbclient).
// v0.4.67: NFS back as in-process Rust client (nfs3_client).
// Both render with the same "remote" badge colour — they
// share the same "on a remote share, can't be deleted from
// here" semantics in the UI.
const isSmb = i.source && i.source.kind === 'smb';
const isNfs = i.source && i.source.kind === 'nfs';
const isRemote = isSmb || isNfs;
const protectedNow = !!i.password_hash;
// The inline editor row is hidden by default; the Password
@@ -587,8 +598,9 @@
]),
el('td', {class:'num'}, fmtBytes(i.size_bytes)),
el('td', {},
el('span', {class:'src-badge' + (isSmb ? ' nfs' : '')},
isSmb ? ('smb:' + i.source.share_id) : 'local')),
el('span', {class:'src-badge' + (isRemote ? ' nfs' : '')},
isSmb ? ('smb:' + i.source.share_id)
: isNfs ? ('nfs:' + i.source.share_id) : 'local')),
el('td', {},
protectedNow
? el('span', {class:'tag accent'}, 'protected')
@@ -598,11 +610,12 @@
el('button', {class:'ghost', style:'margin-right:6px', onclick: () => {
editorRow.style.display = (editorRow.style.display === 'none') ? '' : 'none';
}}, protectedNow ? 'Password ✎' : 'Set password'),
isSmb
// v0.4.65: SMB-sourced ISOs live on the remote share —
// OpenPXE doesn't own those bytes. Same pattern as NFS
// had: surface a tag instead of a destructive button.
? el('span', {class:'tag', style:'opacity:.6'}, 'on SMB')
isRemote
// v0.4.65/v0.4.67: remote-sourced ISOs live on the
// share — OpenPXE doesn't own those bytes. Surface a
// tag instead of a destructive button.
? el('span', {class:'tag', style:'opacity:.6'},
isSmb ? 'on SMB' : 'on NFS')
: el('button', {class:'danger', onclick: async () => {
if (!confirm('Remove this image?')) return;
await fetch('/api/isos/' + encodeURIComponent(i.id), {method:'DELETE'});
@@ -715,6 +728,72 @@
el('span'),
])) : [el('div', {class:'empty'}, 'No SMB shares configured.')];
// ── NFS shares section (v0.4.67) ──
// Parallel to SMB shares above. The NFSv3 client is in-process
// (nfs3_client crate) so NFS-sourced ISOs support HTTP Range
// requests — SMB-sourced ones don't (smbclient CLI can't seek
// mid-stream). Otherwise the UX is identical: server + export,
// submit, scan, remove.
const nfsMsg = el('div', {class:'msg'});
const nfsServerIn = el('input', {type:'text', placeholder:'10.0.0.5'});
const nfsExportIn = el('input', {type:'text', placeholder:'/srv/isos'});
const addNfs = el('button', {onclick: async () => {
if (!nfsServerIn.value || !nfsExportIn.value) {
nfsMsg.replaceChildren(document.createTextNode('Server and export are required.'));
nfsMsg.className = 'msg err'; return;
}
nfsMsg.replaceChildren(document.createTextNode('Connecting…'));
nfsMsg.className = 'msg';
const r = await postJSON('/api/nfs-shares', {
server: nfsServerIn.value,
export: nfsExportIn.value,
});
if (r.ok) {
nfsMsg.replaceChildren(document.createTextNode('Connected.'));
nfsMsg.className = 'msg ok';
render('storage');
} else {
// Structured {error, stderr, hint} same as SMB.
let bodyJson = null;
let raw = null;
try { bodyJson = await r.clone().json(); }
catch (_) { raw = await r.text().catch(()=> 'connect failed'); }
const msg = bodyJson && bodyJson.error ? bodyJson.error : (raw || 'connect failed');
const hint = bodyJson && bodyJson.hint;
const parts = [el('div', {}, [
el('strong', {}, 'Connect failed: '),
document.createTextNode(msg),
])];
if (hint) {
parts.push(el('div', {style:'margin-top:6px;opacity:.78;font-size:12px'}, hint));
}
nfsMsg.replaceChildren(...parts);
nfsMsg.className = 'msg err';
}
}}, 'Add share');
const nfsRows = nfsShares.length ? nfsShares.map(m => el('div', {class: 'nfs-row' + (m.reachable ? '' : ' down')}, [
el('span', {class: 'dot ' + (m.reachable ? 'ok' : 'err')}),
el('div', {}, [
el('div', {class:'id'}, m.server + ':' + m.export),
el('div', {class:'meta'},
'NFSv3 · ' +
(m.reachable ? m.iso_count + ' isos' : 'not reachable')),
m.last_error ? el('div', {class:'err'}, '⚠ ' + m.last_error) : null,
m.last_hint ? el('div', {style:'margin-top:4px;opacity:.78;font-size:12px'}, m.last_hint) : null,
]),
el('button', {class:'ghost', onclick: async () => {
const r = await postJSON('/api/nfs-shares/' + encodeURIComponent(m.id) + '/scan', {});
if (r.ok) render('storage');
}}, 'Re-scan'),
el('button', {class:'danger', onclick: async () => {
if (!confirm('Forget ' + m.server + ':' + m.export + '?')) return;
await fetch('/api/nfs-shares/' + encodeURIComponent(m.id), {method:'DELETE'});
render('storage');
}}, 'Remove'),
el('span'),
])) : [el('div', {class:'empty'}, 'No NFS shares configured.')];
// Disk-space card. Free + used + total for the volume hosting the
// ISO directory, with a coloured bar. Warns at 80% and goes red at
// 95% so the operator sees the runway shrinking before uploads
@@ -801,6 +880,39 @@
'boot time — no local cache, no double disk usage.'),
]),
]),
// v0.4.67: NFS shares card sits right below SMB so operators
// can see both protocols at a glance. The form is simpler
// (no auth) because NFSv3 access control is by client IP on
// the server side, not by client-supplied credentials.
el('div', {class:'card'}, [
el('header', {}, [
el('h2', {}, 'NFS shares'),
el('span', {class:'sub'}, nfsShares.length + ' configured'),
]),
el('div', {class:'body'}, [
el('div', {class:'form-row cols-2'}, [
el('label', {class:'field'}, [
el('span', {class:'name'}, 'NFS server'),
nfsServerIn,
]),
el('label', {class:'field'}, [
el('span', {class:'name'}, 'Export path'),
nfsExportIn,
]),
]),
addNfs, nfsMsg,
el('div', {style:'margin-top:18px;display:grid;gap:8px'}, nfsRows),
el('p', {class:'msg', style:'margin-top:14px'},
'NFSv3 shares are read in-process via a pure-Rust client — ' +
'no kernel modules, no mount.nfs, no CAP_SYS_ADMIN. Works in ' +
'every container the SMB path works in (Unraid included). ' +
'NFSv3 auth is AUTH_SYS only; gate access on the server side ' +
'by allowing this OpenPXE hosts IP in the export list. ' +
'ISOs are streamed on demand and HTTP Range requests work — ' +
'NFSv3 READ3 takes an explicit offset, so clients can seek ' +
'into a 5 GB ISO without reading what comes before.'),
]),
]),
el('div', {class:'card'}, [
el('header', {}, [
el('h2', {}, 'Available images'),