mirror of
https://github.com/scottlamb/moonfire-nvr.git
synced 2026-11-06 16:07:50 +00:00
This puts a lot of state into a new `LockedStream` that was formerly guarded more coarsely by the recording lock or more granularly by `RecordingToInsert` locks. And it restructures `LockedDatabase::flush` to avoid holding this lock while performing SQLite operations. Streamers write only into a recent frames buffer in RAM, which the directory pool workers asynchronously write to disk approximately once per GOP, where formerly there was a write call per frame. Before, streamers could fall behind due to slow sample file directory disk I/O calls or if writing got stuck in a retry loop, for example if out of disk space. If Moonfire falls too far behind in receiving, cameras tend to disconnect or (sadly) mis-frame messages because they do not handle this case well. This removes a confusing potential cause of such errors. In the new model, streamers never wait for disk I/O. Frames are buffered up to a limit. Recordings are aborted if the I/O pool falls too far behind. There's no retry loop. This roughly halves CPU usage as thread hand-offs and syscalls are less common. It also enables several desirable future changes: * audio recording (which radically increases the number of frames and thus would otherwise significantly increase CPU usage). * "monitor mode" (allowing live view without recording to disk), * further CPU reductions for live view (by skipping all SQLite ops and disk I/O). In the new model, streamers also rarely wait for SQLite database operations either—now only on video parameter changes. That might go away in an upcoming change too. The "syncer" (which synced files/dirs and kept track of what has to be flushed into the SQLite database when) has mostly gone away too. The directory pool workers handle the syncing now, and then they notify a new "flusher" for the latter portion. I'd like to go further still and have the database as a whole be owned by a dedicated thread, the only one that uses the SQLite connection. This prevents the tokio threads from blocking on SQLite operations. I tried out doing so before making the streamers less dependent on the db, but the extra thread handoffs were awful for performance.
276 lines
7.7 KiB
Rust
276 lines
7.7 KiB
Rust
// This file is part of Moonfire NVR, a security camera network video recorder.
|
|
// Copyright (C) 2018 The Moonfire NVR Authors; see AUTHORS and LICENSE.txt.
|
|
// SPDX-License-Identifier: GPL-v3.0-or-later WITH GPL-3.0-linking-exception.
|
|
|
|
//! Clock interface and implementations for testability.
|
|
//!
|
|
//! Note these types are in a more standard nanosecond-based format, where
|
|
//! [`crate::time`] uses Moonfire's 90 kHz time base.
|
|
|
|
use crate::Mutex;
|
|
use nix::sys::time::{TimeSpec, TimeValLike as _};
|
|
use std::future::Future;
|
|
use std::panic::Location;
|
|
use std::sync::Arc;
|
|
pub use std::time::Duration;
|
|
use tracing::warn;
|
|
|
|
use crate::error::Error;
|
|
use crate::shutdown::ShutdownError;
|
|
|
|
#[derive(Copy, Clone, PartialEq, Eq)]
|
|
pub struct SystemTime(pub TimeSpec);
|
|
|
|
impl SystemTime {
|
|
pub fn new(sec: u64, nsec: i64) -> Self {
|
|
// `TimeSpec::new`'s arguments vary by platform.
|
|
// * currently uses 32-bit time_t on musl <https://github.com/rust-lang/libc/issues/1848>
|
|
// * nsec likewise can vary.
|
|
SystemTime(TimeSpec::new(sec as _, nsec as _))
|
|
}
|
|
|
|
pub fn as_secs(&self) -> i64 {
|
|
self.0.num_seconds()
|
|
}
|
|
}
|
|
|
|
impl std::ops::Add<Duration> for SystemTime {
|
|
type Output = SystemTime;
|
|
|
|
fn add(self, rhs: Duration) -> SystemTime {
|
|
SystemTime(self.0 + TimeSpec::from(rhs))
|
|
}
|
|
}
|
|
|
|
#[derive(Copy, Clone, PartialEq, Eq, PartialOrd, Ord)]
|
|
pub struct Instant(pub TimeSpec);
|
|
|
|
impl Instant {
|
|
pub fn from_secs(secs: i64) -> Self {
|
|
Instant(TimeSpec::seconds(secs))
|
|
}
|
|
|
|
pub fn saturating_sub(&self, o: &Instant) -> Duration {
|
|
if o > self {
|
|
Duration::default()
|
|
} else {
|
|
Duration::from(self.0 - o.0)
|
|
}
|
|
}
|
|
}
|
|
|
|
impl std::fmt::Debug for Instant {
|
|
fn fmt(&self, f: &mut std::fmt::Formatter<'_>) -> std::fmt::Result {
|
|
self.0.fmt(f)
|
|
}
|
|
}
|
|
|
|
// TODO: should use saturating always?
|
|
impl std::ops::Sub<Instant> for Instant {
|
|
type Output = Duration;
|
|
|
|
fn sub(self, rhs: Instant) -> Duration {
|
|
Duration::from(self.0 - rhs.0)
|
|
}
|
|
}
|
|
|
|
impl std::ops::Add<Duration> for Instant {
|
|
type Output = Instant;
|
|
|
|
fn add(self, rhs: Duration) -> Instant {
|
|
Instant(self.0 + TimeSpec::from(rhs))
|
|
}
|
|
}
|
|
|
|
/// Abstract interface to the system clocks. This is for testability.
|
|
pub trait Clocks: Clone + Send + Sync + 'static {
|
|
/// Gets the current time from `CLOCK_REALTIME`.
|
|
fn realtime(&self) -> SystemTime;
|
|
|
|
/// Gets the current time from a monotonic clock.
|
|
///
|
|
/// On Linux, this uses `CLOCK_BOOTTIME`, which includes suspended time.
|
|
/// On other systems, it uses `CLOCK_MONOTONIC`.
|
|
fn monotonic(&self) -> Instant;
|
|
|
|
/// Causes the current thread to sleep for the specified time.
|
|
fn sleep(&self, how_long: Duration) -> impl Future<Output = ()> + Send;
|
|
}
|
|
|
|
/// Waits a bit before retrying an operation.
|
|
///
|
|
/// Use as follows (for an operation that returns `()` on success):
|
|
///
|
|
/// ```no_compile
|
|
/// while let Err(e) = fallible_operation().await {
|
|
/// retry_wait(clocks, shutdown_rx, e).await?;
|
|
/// }
|
|
/// ```
|
|
///
|
|
/// or (for an operation that returns a value on success):
|
|
///
|
|
/// ```no_compile
|
|
/// let result = loop {
|
|
/// match fallible_operation().await {
|
|
/// Ok(result) => break result,
|
|
/// Err(e) => retry_wait(clocks, shutdown_rx, e).await?,
|
|
/// }
|
|
/// };
|
|
/// ```
|
|
///
|
|
/// # Interface note
|
|
///
|
|
/// The synchronous version of this method took a lambda and encompassed the
|
|
/// retry loop, as follows:
|
|
///
|
|
/// ```no_compile
|
|
/// fn retry<C: Clocks, E: Into<Error>>(
|
|
/// clocks: &C,
|
|
/// shutdown_rx: &crate::shutdown::Receiver,
|
|
/// f: &mut FnMut() -> Result<T, E>,,
|
|
/// ) -> Result<T, ShutdownError> { todo!() }
|
|
/// ```
|
|
///
|
|
/// Unfortunately it does not seem trivial to match this with async as of
|
|
/// Rust 1.85. The following attempt with async closures is close, but
|
|
/// critically the `F(..): Send + 'a` bound is not yet supported. Alternatively,
|
|
/// we could use a `FnMut -> Future<Output = Result<T, E>>` approach, but it's
|
|
/// not possible to allow the future to borrow from the `FnMut`.
|
|
///
|
|
/// ```no_compile
|
|
/// pub fn retry<'a, C, T, E, F>(
|
|
/// clocks: &'a C,
|
|
/// shutdown_rx: &'a crate::shutdown::Receiver,
|
|
/// mut f: F,
|
|
/// ) -> impl Future<Output = Result<T, ShutdownError>> + Send + 'a
|
|
/// where
|
|
/// C: Clocks,
|
|
/// E: Into<Error>,
|
|
/// F: AsyncFnMut() -> Result<T, E> + Send + 'a,
|
|
/// F(..): Send + 'a,
|
|
/// { todo!() }
|
|
/// ```
|
|
///
|
|
/// Alternatively, we could use a `FnMut -> Future<Output = Result<T, E>>`
|
|
/// approach, but it's not possible to allow the future to borrow from the
|
|
/// `FnMut`.
|
|
pub async fn retry_wait<C: Clocks>(
|
|
clocks: &C,
|
|
shutdown_rx: &crate::shutdown::Receiver,
|
|
e: Error,
|
|
) -> Result<(), ShutdownError> {
|
|
warn!(
|
|
exception = %e.chain(),
|
|
"sleeping for 1 s after error"
|
|
);
|
|
tokio::select! {
|
|
biased;
|
|
_ = shutdown_rx.as_future() => Err(ShutdownError),
|
|
_ = clocks.sleep(Duration::from_secs(1)) => Ok(()),
|
|
}
|
|
}
|
|
|
|
#[derive(Copy, Clone)]
|
|
pub struct RealClocks {}
|
|
|
|
impl Clocks for RealClocks {
|
|
fn realtime(&self) -> SystemTime {
|
|
SystemTime(
|
|
nix::time::clock_gettime(nix::time::ClockId::CLOCK_REALTIME)
|
|
.expect("clock_gettime(REALTIME) should succeed"),
|
|
)
|
|
}
|
|
|
|
#[cfg(target_os = "linux")]
|
|
fn monotonic(&self) -> Instant {
|
|
Instant(
|
|
nix::time::clock_gettime(nix::time::ClockId::CLOCK_BOOTTIME)
|
|
.expect("clock_gettime(BOOTTIME) should succeed"),
|
|
)
|
|
}
|
|
|
|
#[cfg(not(target_os = "linux"))]
|
|
fn monotonic(&self) -> Instant {
|
|
Instant(
|
|
nix::time::clock_gettime(nix::time::ClockId::CLOCK_MONOTONIC)
|
|
.expect("clock_gettime(MONOTONIC) should succeed"),
|
|
)
|
|
}
|
|
|
|
fn sleep(&self, how_long: Duration) -> impl Future<Output = ()> + Send {
|
|
tokio::time::sleep(how_long)
|
|
}
|
|
}
|
|
|
|
/// Logs a warning if the TimerGuard lives "too long", using the label created by a supplied
|
|
/// function.
|
|
pub struct TimerGuard<'a, C: Clocks, S: AsRef<str>, F: FnOnce(&'static Location<'static>) -> S + 'a>
|
|
{
|
|
clocks: &'a C,
|
|
location: &'static Location<'static>,
|
|
label_f: Option<F>,
|
|
start: Instant,
|
|
}
|
|
|
|
impl<'a, C: Clocks, S: AsRef<str>, F: FnOnce(&'static Location<'static>) -> S + 'a>
|
|
TimerGuard<'a, C, S, F>
|
|
{
|
|
#[track_caller]
|
|
pub fn new(clocks: &'a C, label_f: F) -> Self {
|
|
TimerGuard {
|
|
clocks,
|
|
location: Location::caller(),
|
|
label_f: Some(label_f),
|
|
start: clocks.monotonic(),
|
|
}
|
|
}
|
|
}
|
|
|
|
impl<'a, C, S, F> Drop for TimerGuard<'a, C, S, F>
|
|
where
|
|
C: Clocks,
|
|
S: AsRef<str>,
|
|
F: FnOnce(&'static Location<'static>) -> S + 'a,
|
|
{
|
|
fn drop(&mut self) {
|
|
let elapsed = self.clocks.monotonic() - self.start;
|
|
if elapsed.as_secs() >= 1 {
|
|
let label_f = self.label_f.take().unwrap();
|
|
warn!("{} took {:?}!", label_f(self.location).as_ref(), elapsed);
|
|
}
|
|
}
|
|
}
|
|
|
|
/// Simulated clock for testing.
|
|
#[derive(Clone)]
|
|
pub struct SimulatedClocks(Arc<SimulatedClocksInner>);
|
|
|
|
struct SimulatedClocksInner {
|
|
boot: SystemTime,
|
|
uptime: Mutex<Duration, 3>,
|
|
}
|
|
|
|
impl SimulatedClocks {
|
|
pub fn new(boot: SystemTime) -> Self {
|
|
SimulatedClocks(Arc::new(SimulatedClocksInner {
|
|
boot,
|
|
uptime: Mutex::new(Duration::from_secs(0)),
|
|
}))
|
|
}
|
|
}
|
|
|
|
impl Clocks for SimulatedClocks {
|
|
fn realtime(&self) -> SystemTime {
|
|
self.0.boot + *self.0.uptime.lock()
|
|
}
|
|
fn monotonic(&self) -> Instant {
|
|
Instant(TimeSpec::from(*self.0.uptime.lock()))
|
|
}
|
|
|
|
/// Advances the clock by the specified amount without actually sleeping.
|
|
async fn sleep(&self, how_long: Duration) {
|
|
let mut l = self.0.uptime.lock();
|
|
*l += how_long;
|
|
}
|
|
}
|