]> git.bitcoin.ninja Git - rust-lightning/commitdiff
Persist full monitor if there is an error while applying monitor_update
authorGursharan Singh <3442979+G8XSU@users.noreply.github.com>
Fri, 29 Sep 2023 03:30:53 +0000 (20:30 -0700)
committerGursharan Singh <3442979+G8XSU@users.noreply.github.com>
Fri, 29 Sep 2023 23:26:40 +0000 (16:26 -0700)
Motivation: When there is an error while applying monitor_update to a
channel_monitor,  we don't want to persist a 'monitor_update' which
results in a failure to apply later while reading 'channel_monitor' with
updates from storage. Instead, we should persist the entire 'channel_monitor'
here.

lightning/src/chain/chainmonitor.rs

index 81d23b3ae2079a5215beffac205f4f09dc9bd87e..086a848af77aa05c7ea4e8e97e93c4e6fb5e9628 100644 (file)
@@ -764,14 +764,20 @@ where C::Target: chain::Filter,
                                let monitor = &monitor_state.monitor;
                                log_trace!(self.logger, "Updating ChannelMonitor for channel {}", log_funding_info!(monitor));
                                let update_res = monitor.update_monitor(update, &self.broadcaster, &*self.fee_estimator, &self.logger);
-                               if update_res.is_err() {
-                                       log_error!(self.logger, "Failed to update ChannelMonitor for channel {}.", log_funding_info!(monitor));
-                               }
-                               // Even if updating the monitor returns an error, the monitor's state will
-                               // still be changed. So, persist the updated monitor despite the error.
+
                                let update_id = MonitorUpdateId::from_monitor_update(update);
                                let mut pending_monitor_updates = monitor_state.pending_monitor_updates.lock().unwrap();
-                               let persist_res = self.persister.update_persisted_channel(funding_txo, Some(update), monitor, update_id);
+                               let persist_res = if update_res.is_err() {
+                                       // Even if updating the monitor returns an error, the monitor's state will
+                                       // still be changed. Therefore, we should persist the updated monitor despite the error.
+                                       // We don't want to persist a `monitor_update` which results in a failure to apply later
+                                       // while reading `channel_monitor` with updates from storage. Instead, we should persist
+                                       // the entire `channel_monitor` here.
+                                       log_warn!(self.logger, "Failed to update ChannelMonitor for channel {}. Going ahead and persisting the entire ChannelMonitor", log_funding_info!(monitor));
+                                       self.persister.update_persisted_channel(funding_txo, None, monitor, update_id)
+                               } else {
+                                       self.persister.update_persisted_channel(funding_txo, Some(update), monitor, update_id)
+                               };
                                match persist_res {
                                        ChannelMonitorUpdateStatus::InProgress => {
                                                pending_monitor_updates.push(update_id);