nostr/nostr-rs-relay
1
0
Fork 0
mirror of https://github.com/scsibug/nostr-rs-relay.git synced 2024-11-22 00:59:07 -05:00
Code Issues Packages Projects Releases Wiki Activity

fix: use database to publish all events

Browse Source 
This fixes a race condition where a publisher might send an event, and
immediately after issue a subscription for the same event ID.  Prior
to this change, that event would have been published on the broadcast
channel (and ignored by our publisher, because they had not yet issued
the subscription), but not yet committed to the database.  Their
subscription would trigger a database query which would return zero
results.  Therefore, they would never see the event they published.
The noscl tool is one client that would suffer from this.

Now, all events are broadcast only after they exist in the database,
so a late subscription will always return the event.
This commit is contained in:
Greg Heartsfield 2021-12-12 10:20:23 -06:00
parent 56c40f2be9
commit 1589268eba
2 changed files with 7 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
src/db.rs
View File

@ -67,6 +67,7 @@ CREATE INDEX IF NOT EXISTS pubkey_ref_index ON pubkey_ref(referenced_pubkey);
/// Spawn a database writer that persists events to the SQLite store.
pub async fn db_writer(
mut event_rx: tokio::sync::mpsc::Receiver<Event>,
bcast_tx: tokio::sync::broadcast::Sender<Event>,
) -> tokio::task::JoinHandle<Result<()>> {
task::spawn_blocking(move || {
let mut conn = Connection::open_with_flags(
@ -94,6 +95,8 @@ pub async fn db_writer(
info!("nothing inserted (dupe?)");
} else {
info!("persisted event: {}", event.get_event_id_prefix());
// send this out to all clients
bcast_tx.send(event.clone()).ok();
}
}
Err(err) => {

13
src/main.rs
View File

@ -44,8 +44,10 @@ fn main() -> Result<(), Error> {
// validated events that need to be persisted are sent to the
// database on via this channel.
let (event_tx, event_rx) = mpsc::channel::<Event>(16);
// start the database writer thread.
db::db_writer(event_rx).await;
// start the database writer thread. Give it a channel for
// writing events, and for publishing events that have been
// written (to all connected clients).
db::db_writer(event_rx, bcast_tx.clone()).await;
// establish a channel for letting all threads now about a
// requested server shutdown.
let (invoke_shutdown, _) = broadcast::channel::<()>(1);
@ -86,7 +88,6 @@ async fn nostr_server(
mut shutdown: Receiver<()>,
) {
// get a broadcast channel for clients to communicate on
// wrap the TCP stream in a websocket.
let mut bcast_rx = broadcast.subscribe();
// upgrade the TCP connection to WebSocket
let conn = tokio_tungstenite::accept_async(stream).await;
@ -154,18 +155,12 @@ async fn nostr_server(
// Write this to the database
event_tx.send(e.clone()).await.ok();
client_published_event_count += 1;
// send this event to everyone listening.
let bcast_res = broadcast.send(e);
if bcast_res.is_err() {
warn!("could not send broadcast message: {:?}", bcast_res);
}
},
Err(_) => {info!("client {} sent an invalid event", cid)}
}
},
Some(Ok(SubMsg(s))) => {
debug!("client {} requesting a subscription", cid);
// subscription handling consists of:
// * registering the subscription so future events can be matched
// * making a channel to cancel to request later