Add db queries

This commit is contained in:
Edward Shen 2021-04-22 21:21:32 -04:00
parent b01fa618b4
commit 288872e84b
Signed by: edward
GPG key ID: 19182661E818369F
5 changed files with 104 additions and 29 deletions

View file

@ -1,7 +1,6 @@
create table if not exists Images(
id varchar primary key not null,
size integer not null,
accessed timestamp not null default CURRENT_TIMESTAMP,
disk_size integer as ((size + 4095) / 4096 * 4096)
accessed timestamp not null default CURRENT_TIMESTAMP
);
create index if not exists Images_accessed on Images(accessed);

View file

@ -6,4 +6,5 @@
# We can trust that our program will initialize the db at runtime the same way
# as it pulls from the same file for initialization
mkdir cache
sqlite3 cache/metadata.sqlite < db_queries/init.sql

28
src/cache/fs.rs vendored
View file

@ -1,13 +1,13 @@
use actix_web::error::PayloadError;
use bytes::{Buf, Bytes};
use futures::{Stream, StreamExt};
use futures::{Future, Stream, StreamExt};
use log::{debug, error};
use once_cell::sync::Lazy;
use serde::Deserialize;
use std::collections::HashMap;
use std::fmt::Display;
use std::io::SeekFrom;
use std::num::NonZeroU64;
use std::num::NonZeroU32;
use std::path::{Path, PathBuf};
use std::pin::Pin;
use std::task::{Context, Poll};
@ -69,11 +69,15 @@ pub async fn read_file(
/// Maps the input byte stream into one that writes to disk instead, returning
/// a stream that reads from disk instead.
pub async fn write_file(
pub async fn write_file<
Fut: 'static + Send + Sync + Future<Output = ()>,
F: 'static + Send + Sync + FnOnce(u32) -> Fut,
>(
path: &Path,
mut byte_stream: BoxedImageStream,
metadata: ImageMetadata,
notifier: UnboundedSender<u64>,
notifier: UnboundedSender<u32>,
db_callback: F,
) -> Result<CacheStream, std::io::Error> {
let (tx, rx) = channel(WritingStatus::NotDone);
@ -93,7 +97,7 @@ pub async fn write_file(
tokio::spawn(async move {
let path_buf = path_buf; // moves path buf into async
let mut errored = false;
let mut bytes_written: u64 = 0;
let mut bytes_written: u32 = 0;
file.write_all(&metadata.as_bytes()).await?;
while let Some(bytes) = byte_stream.next().await {
if let Ok(mut bytes) = bytes {
@ -103,7 +107,7 @@ pub async fn write_file(
n => {
bytes.advance(n);
// We don't care if we don't have receivers
bytes_written += n as u64;
bytes_written += n as u32;
let _ = tx.send(WritingStatus::NotDone);
}
}
@ -146,6 +150,8 @@ pub async fn write_file(
);
}
tokio::spawn(db_callback(bytes_written));
// We don't ever check this, so the return value doesn't matter
Ok::<_, std::io::Error>(())
});
@ -158,8 +164,8 @@ pub async fn write_file(
pub struct ConcurrentFsStream {
file: Pin<Box<BufReader<File>>>,
receiver: Pin<Box<WatchStream<WritingStatus>>>,
bytes_read: u64,
bytes_total: Option<NonZeroU64>,
bytes_read: u32,
bytes_total: Option<NonZeroU32>,
}
impl ConcurrentFsStream {
@ -225,7 +231,7 @@ impl Stream for ConcurrentFsStream {
if let Poll::Ready(Some(WritingStatus::Done(n))) =
self.receiver.as_mut().poll_next_unpin(cx)
{
self.bytes_total = Some(NonZeroU64::new(n).unwrap())
self.bytes_total = Some(NonZeroU32::new(n).unwrap())
}
// Okay, now we know if we've read enough bytes or not. If the
@ -248,7 +254,7 @@ impl Stream for ConcurrentFsStream {
Poll::Ready(Some(Ok(Bytes::new())))
} else {
// We have data! Give it to the reader!
self.bytes_read += filled as u64;
self.bytes_read += filled as u32;
bytes.truncate(filled);
Poll::Ready(Some(Ok(bytes.into())))
}
@ -265,6 +271,6 @@ impl From<UpstreamError> for actix_web::Error {
#[derive(Debug, Clone, Copy)]
enum WritingStatus {
NotDone,
Done(u64),
Done(u32),
Error,
}

95
src/cache/low_mem.rs vendored
View file

@ -1,29 +1,32 @@
//! Low memory caching stuff
use std::path::PathBuf;
use std::str::FromStr;
use std::sync::{atomic::Ordering, Arc};
use std::{path::PathBuf, sync::atomic::AtomicU64};
use std::sync::atomic::{AtomicU64, Ordering};
use std::sync::Arc;
use async_trait::async_trait;
use futures::StreamExt;
use log::LevelFilter;
use log::{warn, LevelFilter};
use sqlx::{sqlite::SqliteConnectOptions, ConnectOptions, SqlitePool};
use tokio::sync::mpsc::{channel, unbounded_channel, Sender, UnboundedSender};
use tokio::{
fs::remove_file,
sync::mpsc::{channel, unbounded_channel, Sender, UnboundedSender},
};
use tokio_stream::wrappers::ReceiverStream;
use super::{BoxedImageStream, Cache, CacheError, CacheKey, CacheStream, ImageMetadata};
pub struct LowMemCache {
disk_path: PathBuf,
disk_max_size: u64,
disk_cur_size: AtomicU64,
file_size_channel_sender: UnboundedSender<u64>,
file_size_channel_sender: UnboundedSender<u32>,
db_update_channel_sender: Sender<DbMessage>,
}
enum DbMessage {
Get(Arc<CacheKey>),
Put(Arc<CacheKey>, ImageMetadata),
Put(PathBuf, u32),
}
impl LowMemCache {
@ -53,7 +56,6 @@ impl LowMemCache {
let new_self: Arc<Box<dyn Cache>> = Arc::new(Box::new(Self {
disk_path,
disk_max_size,
disk_cur_size: AtomicU64::new(0),
file_size_channel_sender: file_size_tx,
db_update_channel_sender: db_tx,
@ -63,10 +65,30 @@ impl LowMemCache {
// the channel, which informs the low memory cache the total size of the
// item that was put into the cache.
let new_self_0 = Arc::clone(&new_self);
let db_pool_0 = db_pool.clone();
tokio::spawn(async move {
let db_pool = db_pool_0;
let max_on_disk_size = disk_max_size / 20 * 19;
// This will never return None, effectively a loop
while let Some(new_size) = file_size_rx.recv().await {
new_self_0.increase_usage(new_size).await;
new_self_0.increase_usage(new_size);
if new_self_0.on_disk_size() >= max_on_disk_size {
let mut conn = db_pool.acquire().await.unwrap();
let items = sqlx::query!(
"select id, size from Images order by accessed asc limit 1000"
)
.fetch_all(&mut conn)
.await
.unwrap();
let mut size_freed = 0;
for item in items {
size_freed += item.size as u64;
tokio::spawn(remove_file(item.id));
}
new_self_0.decrease_usage(size_freed);
}
}
});
@ -75,9 +97,39 @@ impl LowMemCache {
let db_pool = db_pool;
let mut recv_stream = ReceiverStream::new(db_rx).ready_chunks(128);
while let Some(messages) = recv_stream.next().await {
let now = chrono::Utc::now();
let mut transaction = db_pool.begin().await.unwrap();
for message in messages {}
for message in messages {
match message {
DbMessage::Get(entry) => {
let key = entry.to_string();
let query = sqlx::query!(
"update Images set accessed = ? where id = ?",
now,
key
)
.execute(&mut transaction)
.await;
if let Err(e) = query {
warn!("Failed to update timestamp in db for {}: {}", key, e);
}
}
DbMessage::Put(entry, size) => {
let key = entry.as_os_str().to_str();
let query = sqlx::query!(
"insert into Images (id, size, accessed) values (?, ?, ?)",
key,
size,
now,
)
.execute(&mut transaction)
.await;
if let Err(e) = query {
warn!("Failed to add {:?} to db: {}", key, e);
}
}
}
}
transaction.commit().await.unwrap();
}
});
@ -113,22 +165,35 @@ impl Cache for LowMemCache {
metadata: ImageMetadata,
) -> Result<CacheStream, CacheError> {
let channel = self.db_update_channel_sender.clone();
let key_0 = Arc::clone(&key);
tokio::spawn(async move { channel.send(DbMessage::Put(key_0, metadata)).await });
let path = self.disk_path.clone().join(PathBuf::from(key.as_ref()));
let path_0 = path.clone();
let db_callback = |size: u32| async move {
let _ = channel.send(DbMessage::Put(path_0, size)).await;
};
super::fs::write_file(
&path,
image,
metadata,
self.file_size_channel_sender.clone(),
db_callback,
)
.await
.map_err(Into::into)
}
async fn increase_usage(&self, amt: u64) {
self.disk_cur_size.fetch_add(amt, Ordering::Release);
#[inline]
fn increase_usage(&self, amt: u32) {
self.disk_cur_size.fetch_add(amt as u64, Ordering::Release);
}
#[inline]
fn on_disk_size(&self) -> u64 {
(self.disk_cur_size.load(Ordering::Acquire) + 4095) / 4096 * 4096
}
fn decrease_usage(&self, amt: u64) {
self.disk_cur_size.fetch_sub(amt, Ordering::Release);
}
}

6
src/cache/mod.rs vendored
View file

@ -164,7 +164,11 @@ pub trait Cache: Send + Sync {
metadata: ImageMetadata,
) -> Result<CacheStream, CacheError>;
async fn increase_usage(&self, amt: u64);
fn increase_usage(&self, amt: u32);
fn decrease_usage(&self, amt: u64);
fn on_disk_size(&self) -> u64;
}
pub enum CacheStream {