1 // This file is licensed under the Apache License, Version 2.0 <LICENSE-APACHE
2 // or http://www.apache.org/licenses/LICENSE-2.0> or the MIT license
3 // <LICENSE-MIT or http://opensource.org/licenses/MIT>, at your option.
4 // You may not use this file except in accordance with one or both of these
7 //! This module contains a simple key-value store trait [`KVStore`] that
8 //! allows one to implement the persistence for [`ChannelManager`], [`NetworkGraph`],
9 //! and [`ChannelMonitor`] all in one place.
11 //! [`ChannelManager`]: crate::ln::channelmanager::ChannelManager
15 use core::str::FromStr;
16 use bitcoin::{BlockHash, Txid};
18 use crate::{io, log_error};
19 use crate::prelude::*;
22 use crate::chain::chaininterface::{BroadcasterInterface, FeeEstimator};
23 use crate::chain::chainmonitor::{Persist, MonitorUpdateId};
24 use crate::sign::{EntropySource, ecdsa::WriteableEcdsaChannelSigner, SignerProvider};
25 use crate::chain::transaction::OutPoint;
26 use crate::chain::channelmonitor::{ChannelMonitor, ChannelMonitorUpdate, CLOSED_CHANNEL_UPDATE_ID};
27 use crate::ln::channelmanager::AChannelManager;
28 use crate::routing::gossip::NetworkGraph;
29 use crate::routing::scoring::WriteableScore;
30 use crate::util::logger::Logger;
31 use crate::util::ser::{Readable, ReadableArgs, Writeable};
33 /// The alphabet of characters allowed for namespaces and keys.
34 pub const KVSTORE_NAMESPACE_KEY_ALPHABET: &str = "abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789_-";
36 /// The maximum number of characters namespaces and keys may have.
37 pub const KVSTORE_NAMESPACE_KEY_MAX_LEN: usize = 120;
39 /// The primary namespace under which the [`ChannelManager`] will be persisted.
41 /// [`ChannelManager`]: crate::ln::channelmanager::ChannelManager
42 pub const CHANNEL_MANAGER_PERSISTENCE_PRIMARY_NAMESPACE: &str = "";
43 /// The secondary namespace under which the [`ChannelManager`] will be persisted.
45 /// [`ChannelManager`]: crate::ln::channelmanager::ChannelManager
46 pub const CHANNEL_MANAGER_PERSISTENCE_SECONDARY_NAMESPACE: &str = "";
47 /// The key under which the [`ChannelManager`] will be persisted.
49 /// [`ChannelManager`]: crate::ln::channelmanager::ChannelManager
50 pub const CHANNEL_MANAGER_PERSISTENCE_KEY: &str = "manager";
52 /// The primary namespace under which [`ChannelMonitor`]s will be persisted.
53 pub const CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE: &str = "monitors";
54 /// The secondary namespace under which [`ChannelMonitor`]s will be persisted.
55 pub const CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE: &str = "";
56 /// The primary namespace under which [`ChannelMonitorUpdate`]s will be persisted.
57 pub const CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE: &str = "monitor_updates";
59 /// The primary namespace under which archived [`ChannelMonitor`]s will be persisted.
60 pub const ARCHIVED_CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE: &str = "archived_monitors";
61 /// The secondary namespace under which archived [`ChannelMonitor`]s will be persisted.
62 pub const ARCHIVED_CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE: &str = "";
64 /// The primary namespace under which the [`NetworkGraph`] will be persisted.
65 pub const NETWORK_GRAPH_PERSISTENCE_PRIMARY_NAMESPACE: &str = "";
66 /// The secondary namespace under which the [`NetworkGraph`] will be persisted.
67 pub const NETWORK_GRAPH_PERSISTENCE_SECONDARY_NAMESPACE: &str = "";
68 /// The key under which the [`NetworkGraph`] will be persisted.
69 pub const NETWORK_GRAPH_PERSISTENCE_KEY: &str = "network_graph";
71 /// The primary namespace under which the [`WriteableScore`] will be persisted.
72 pub const SCORER_PERSISTENCE_PRIMARY_NAMESPACE: &str = "";
73 /// The secondary namespace under which the [`WriteableScore`] will be persisted.
74 pub const SCORER_PERSISTENCE_SECONDARY_NAMESPACE: &str = "";
75 /// The key under which the [`WriteableScore`] will be persisted.
76 pub const SCORER_PERSISTENCE_KEY: &str = "scorer";
78 /// A sentinel value to be prepended to monitors persisted by the [`MonitorUpdatingPersister`].
80 /// This serves to prevent someone from accidentally loading such monitors (which may need
81 /// updates applied to be current) with another implementation.
82 pub const MONITOR_UPDATING_PERSISTER_PREPEND_SENTINEL: &[u8] = &[0xFF; 2];
84 /// Provides an interface that allows storage and retrieval of persisted values that are associated
87 /// In order to avoid collisions the key space is segmented based on the given `primary_namespace`s
88 /// and `secondary_namespace`s. Implementations of this trait are free to handle them in different
89 /// ways, as long as per-namespace key uniqueness is asserted.
91 /// Keys and namespaces are required to be valid ASCII strings in the range of
92 /// [`KVSTORE_NAMESPACE_KEY_ALPHABET`] and no longer than [`KVSTORE_NAMESPACE_KEY_MAX_LEN`]. Empty
93 /// primary namespaces and secondary namespaces (`""`) are assumed to be a valid, however, if
94 /// `primary_namespace` is empty, `secondary_namespace` is required to be empty, too. This means
95 /// that concerns should always be separated by primary namespace first, before secondary
96 /// namespaces are used. While the number of primary namespaces will be relatively small and is
97 /// determined at compile time, there may be many secondary namespaces per primary namespace. Note
98 /// that per-namespace uniqueness needs to also hold for keys *and* namespaces in any given
99 /// namespace, i.e., conflicts between keys and equally named
100 /// primary namespaces/secondary namespaces must be avoided.
102 /// **Note:** Users migrating custom persistence backends from the pre-v0.0.117 `KVStorePersister`
103 /// interface can use a concatenation of `[{primary_namespace}/[{secondary_namespace}/]]{key}` to
104 /// recover a `key` compatible with the data model previously assumed by `KVStorePersister::persist`.
106 /// Returns the data stored for the given `primary_namespace`, `secondary_namespace`, and
109 /// Returns an [`ErrorKind::NotFound`] if the given `key` could not be found in the given
110 /// `primary_namespace` and `secondary_namespace`.
112 /// [`ErrorKind::NotFound`]: io::ErrorKind::NotFound
113 fn read(&self, primary_namespace: &str, secondary_namespace: &str, key: &str) -> Result<Vec<u8>, io::Error>;
114 /// Persists the given data under the given `key`.
116 /// Will create the given `primary_namespace` and `secondary_namespace` if not already present
118 fn write(&self, primary_namespace: &str, secondary_namespace: &str, key: &str, buf: &[u8]) -> Result<(), io::Error>;
119 /// Removes any data that had previously been persisted under the given `key`.
121 /// If the `lazy` flag is set to `true`, the backend implementation might choose to lazily
122 /// remove the given `key` at some point in time after the method returns, e.g., as part of an
123 /// eventual batch deletion of multiple keys. As a consequence, subsequent calls to
124 /// [`KVStore::list`] might include the removed key until the changes are actually persisted.
126 /// Note that while setting the `lazy` flag reduces the I/O burden of multiple subsequent
127 /// `remove` calls, it also influences the atomicity guarantees as lazy `remove`s could
128 /// potentially get lost on crash after the method returns. Therefore, this flag should only be
129 /// set for `remove` operations that can be safely replayed at a later time.
131 /// Returns successfully if no data will be stored for the given `primary_namespace`,
132 /// `secondary_namespace`, and `key`, independently of whether it was present before its
133 /// invokation or not.
134 fn remove(&self, primary_namespace: &str, secondary_namespace: &str, key: &str, lazy: bool) -> Result<(), io::Error>;
135 /// Returns a list of keys that are stored under the given `secondary_namespace` in
136 /// `primary_namespace`.
138 /// Returns the keys in arbitrary order, so users requiring a particular order need to sort the
139 /// returned keys. Returns an empty list if `primary_namespace` or `secondary_namespace` is unknown.
140 fn list(&self, primary_namespace: &str, secondary_namespace: &str) -> Result<Vec<String>, io::Error>;
143 /// Trait that handles persisting a [`ChannelManager`], [`NetworkGraph`], and [`WriteableScore`] to disk.
145 /// [`ChannelManager`]: crate::ln::channelmanager::ChannelManager
146 pub trait Persister<'a, CM: Deref, L: Deref, S: WriteableScore<'a>>
148 CM::Target: 'static + AChannelManager,
149 L::Target: 'static + Logger,
151 /// Persist the given ['ChannelManager'] to disk, returning an error if persistence failed.
153 /// [`ChannelManager`]: crate::ln::channelmanager::ChannelManager
154 fn persist_manager(&self, channel_manager: &CM) -> Result<(), io::Error>;
156 /// Persist the given [`NetworkGraph`] to disk, returning an error if persistence failed.
157 fn persist_graph(&self, network_graph: &NetworkGraph<L>) -> Result<(), io::Error>;
159 /// Persist the given [`WriteableScore`] to disk, returning an error if persistence failed.
160 fn persist_scorer(&self, scorer: &S) -> Result<(), io::Error>;
164 impl<'a, A: KVStore + ?Sized, CM: Deref, L: Deref, S: WriteableScore<'a>> Persister<'a, CM, L, S> for A
166 CM::Target: 'static + AChannelManager,
167 L::Target: 'static + Logger,
169 fn persist_manager(&self, channel_manager: &CM) -> Result<(), io::Error> {
170 self.write(CHANNEL_MANAGER_PERSISTENCE_PRIMARY_NAMESPACE,
171 CHANNEL_MANAGER_PERSISTENCE_SECONDARY_NAMESPACE,
172 CHANNEL_MANAGER_PERSISTENCE_KEY,
173 &channel_manager.get_cm().encode())
176 fn persist_graph(&self, network_graph: &NetworkGraph<L>) -> Result<(), io::Error> {
177 self.write(NETWORK_GRAPH_PERSISTENCE_PRIMARY_NAMESPACE,
178 NETWORK_GRAPH_PERSISTENCE_SECONDARY_NAMESPACE,
179 NETWORK_GRAPH_PERSISTENCE_KEY,
180 &network_graph.encode())
183 fn persist_scorer(&self, scorer: &S) -> Result<(), io::Error> {
184 self.write(SCORER_PERSISTENCE_PRIMARY_NAMESPACE,
185 SCORER_PERSISTENCE_SECONDARY_NAMESPACE,
186 SCORER_PERSISTENCE_KEY,
191 impl<ChannelSigner: WriteableEcdsaChannelSigner, K: KVStore + ?Sized> Persist<ChannelSigner> for K {
192 // TODO: We really need a way for the persister to inform the user that its time to crash/shut
193 // down once these start returning failure.
194 // Then we should return InProgress rather than UnrecoverableError, implying we should probably
195 // just shut down the node since we're not retrying persistence!
197 fn persist_new_channel(&self, funding_txo: OutPoint, monitor: &ChannelMonitor<ChannelSigner>, _update_id: MonitorUpdateId) -> chain::ChannelMonitorUpdateStatus {
198 let key = format!("{}_{}", funding_txo.txid.to_string(), funding_txo.index);
200 CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE,
201 CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE,
202 &key, &monitor.encode())
204 Ok(()) => chain::ChannelMonitorUpdateStatus::Completed,
205 Err(_) => chain::ChannelMonitorUpdateStatus::UnrecoverableError
209 fn update_persisted_channel(&self, funding_txo: OutPoint, _update: Option<&ChannelMonitorUpdate>, monitor: &ChannelMonitor<ChannelSigner>, _update_id: MonitorUpdateId) -> chain::ChannelMonitorUpdateStatus {
210 let key = format!("{}_{}", funding_txo.txid.to_string(), funding_txo.index);
212 CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE,
213 CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE,
214 &key, &monitor.encode())
216 Ok(()) => chain::ChannelMonitorUpdateStatus::Completed,
217 Err(_) => chain::ChannelMonitorUpdateStatus::UnrecoverableError
221 fn archive_persisted_channel(&self, funding_txo: OutPoint) {
222 let monitor_name = MonitorName::from(funding_txo);
223 let monitor = match self.read(
224 CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE,
225 CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE,
226 monitor_name.as_str(),
228 Ok(monitor) => monitor,
232 ARCHIVED_CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE,
233 ARCHIVED_CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE,
234 monitor_name.as_str(),
241 CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE,
242 CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE,
243 monitor_name.as_str(),
249 /// Read previously persisted [`ChannelMonitor`]s from the store.
250 pub fn read_channel_monitors<K: Deref, ES: Deref, SP: Deref>(
251 kv_store: K, entropy_source: ES, signer_provider: SP,
252 ) -> Result<Vec<(BlockHash, ChannelMonitor<<SP::Target as SignerProvider>::EcdsaSigner>)>, io::Error>
255 ES::Target: EntropySource + Sized,
256 SP::Target: SignerProvider + Sized,
258 let mut res = Vec::new();
260 for stored_key in kv_store.list(
261 CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE, CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE)?
263 if stored_key.len() < 66 {
264 return Err(io::Error::new(
265 io::ErrorKind::InvalidData,
266 "Stored key has invalid length"));
269 let txid = Txid::from_str(stored_key.split_at(64).0).map_err(|_| {
270 io::Error::new(io::ErrorKind::InvalidData, "Invalid tx ID in stored key")
273 let index: u16 = stored_key.split_at(65).1.parse().map_err(|_| {
274 io::Error::new(io::ErrorKind::InvalidData, "Invalid tx index in stored key")
277 match <(BlockHash, ChannelMonitor<<SP::Target as SignerProvider>::EcdsaSigner>)>::read(
278 &mut io::Cursor::new(
279 kv_store.read(CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE, CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE, &stored_key)?),
280 (&*entropy_source, &*signer_provider),
282 Ok((block_hash, channel_monitor)) => {
283 if channel_monitor.get_funding_txo().0.txid != txid
284 || channel_monitor.get_funding_txo().0.index != index
286 return Err(io::Error::new(
287 io::ErrorKind::InvalidData,
288 "ChannelMonitor was stored under the wrong key",
291 res.push((block_hash, channel_monitor));
294 return Err(io::Error::new(
295 io::ErrorKind::InvalidData,
296 "Failed to read ChannelMonitor"
304 /// Implements [`Persist`] in a way that writes and reads both [`ChannelMonitor`]s and
305 /// [`ChannelMonitorUpdate`]s.
309 /// The main benefit this provides over the [`KVStore`]'s [`Persist`] implementation is decreased
310 /// I/O bandwidth and storage churn, at the expense of more IOPS (including listing, reading, and
311 /// deleting) and complexity. This is because it writes channel monitor differential updates,
312 /// whereas the other (default) implementation rewrites the entire monitor on each update. For
313 /// routing nodes, updates can happen many times per second to a channel, and monitors can be tens
314 /// of megabytes (or more). Updates can be as small as a few hundred bytes.
316 /// Note that monitors written with `MonitorUpdatingPersister` are _not_ backward-compatible with
317 /// the default [`KVStore`]'s [`Persist`] implementation. They have a prepended byte sequence,
318 /// [`MONITOR_UPDATING_PERSISTER_PREPEND_SENTINEL`], applied to prevent deserialization with other
319 /// persisters. This is because monitors written by this struct _may_ have unapplied updates. In
320 /// order to downgrade, you must ensure that all updates are applied to the monitor, and remove the
323 /// # Storing monitors
325 /// Monitors are stored by implementing the [`Persist`] trait, which has two functions:
327 /// - [`Persist::persist_new_channel`], which persists whole [`ChannelMonitor`]s.
328 /// - [`Persist::update_persisted_channel`], which persists only a [`ChannelMonitorUpdate`]
330 /// Whole [`ChannelMonitor`]s are stored in the [`CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE`],
331 /// using the familiar encoding of an [`OutPoint`] (for example, `[SOME-64-CHAR-HEX-STRING]_1`).
333 /// Each [`ChannelMonitorUpdate`] is stored in a dynamic secondary namespace, as follows:
335 /// - primary namespace: [`CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE`]
336 /// - secondary namespace: [the monitor's encoded outpoint name]
338 /// Under that secondary namespace, each update is stored with a number string, like `21`, which
339 /// represents its `update_id` value.
341 /// For example, consider this channel, named for its transaction ID and index, or [`OutPoint`]:
343 /// - Transaction ID: `deadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeef`
346 /// Full channel monitors would be stored at a single key:
348 /// `[CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE]/deadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeef_1`
350 /// Updates would be stored as follows (with `/` delimiting primary_namespace/secondary_namespace/key):
353 /// [CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE]/deadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeef_1/1
354 /// [CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE]/deadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeef_1/2
355 /// [CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE]/deadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeef_1/3
359 /// # Reading channel state from storage
361 /// Channel state can be reconstructed by calling
362 /// [`MonitorUpdatingPersister::read_all_channel_monitors_with_updates`]. Alternatively, users can
363 /// list channel monitors themselves and load channels individually using
364 /// [`MonitorUpdatingPersister::read_channel_monitor_with_updates`].
366 /// ## EXTREMELY IMPORTANT
368 /// It is extremely important that your [`KVStore::read`] implementation uses the
369 /// [`io::ErrorKind::NotFound`] variant correctly: that is, when a file is not found, and _only_ in
370 /// that circumstance (not when there is really a permissions error, for example). This is because
371 /// neither channel monitor reading function lists updates. Instead, either reads the monitor, and
372 /// using its stored `update_id`, synthesizes update storage keys, and tries them in sequence until
373 /// one is not found. All _other_ errors will be bubbled up in the function's [`Result`].
375 /// # Pruning stale channel updates
377 /// Stale updates are pruned when the consolidation threshold is reached according to `maximum_pending_updates`.
378 /// Monitor updates in the range between the latest `update_id` and `update_id - maximum_pending_updates`
380 /// The `lazy` flag is used on the [`KVStore::remove`] method, so there are no guarantees that the deletions
381 /// will complete. However, stale updates are not a problem for data integrity, since updates are
382 /// only read that are higher than the stored [`ChannelMonitor`]'s `update_id`.
384 /// If you have many stale updates stored (such as after a crash with pending lazy deletes), and
385 /// would like to get rid of them, consider using the
386 /// [`MonitorUpdatingPersister::cleanup_stale_updates`] function.
387 pub struct MonitorUpdatingPersister<K: Deref, L: Deref, ES: Deref, SP: Deref>
391 ES::Target: EntropySource + Sized,
392 SP::Target: SignerProvider + Sized,
396 maximum_pending_updates: u64,
402 impl<K: Deref, L: Deref, ES: Deref, SP: Deref>
403 MonitorUpdatingPersister<K, L, ES, SP>
407 ES::Target: EntropySource + Sized,
408 SP::Target: SignerProvider + Sized,
410 /// Constructs a new [`MonitorUpdatingPersister`].
412 /// The `maximum_pending_updates` parameter controls how many updates may be stored before a
413 /// [`MonitorUpdatingPersister`] consolidates updates by writing a full monitor. Note that
414 /// consolidation will frequently occur with fewer updates than what you set here; this number
415 /// is merely the maximum that may be stored. When setting this value, consider that for higher
416 /// values of `maximum_pending_updates`:
418 /// - [`MonitorUpdatingPersister`] will tend to write more [`ChannelMonitorUpdate`]s than
419 /// [`ChannelMonitor`]s, approaching one [`ChannelMonitor`] write for every
420 /// `maximum_pending_updates` [`ChannelMonitorUpdate`]s.
421 /// - [`MonitorUpdatingPersister`] will issue deletes differently. Lazy deletes will come in
422 /// "waves" for each [`ChannelMonitor`] write. A larger `maximum_pending_updates` means bigger,
423 /// less frequent "waves."
424 /// - [`MonitorUpdatingPersister`] will potentially have more listing to do if you need to run
425 /// [`MonitorUpdatingPersister::cleanup_stale_updates`].
427 kv_store: K, logger: L, maximum_pending_updates: u64, entropy_source: ES,
430 MonitorUpdatingPersister {
433 maximum_pending_updates,
439 /// Reads all stored channel monitors, along with any stored updates for them.
441 /// It is extremely important that your [`KVStore::read`] implementation uses the
442 /// [`io::ErrorKind::NotFound`] variant correctly. For more information, please see the
443 /// documentation for [`MonitorUpdatingPersister`].
444 pub fn read_all_channel_monitors_with_updates<B: Deref, F: Deref>(
445 &self, broadcaster: &B, fee_estimator: &F,
446 ) -> Result<Vec<(BlockHash, ChannelMonitor<<SP::Target as SignerProvider>::EcdsaSigner>)>, io::Error>
448 B::Target: BroadcasterInterface,
449 F::Target: FeeEstimator,
451 let monitor_list = self.kv_store.list(
452 CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE,
453 CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE,
455 let mut res = Vec::with_capacity(monitor_list.len());
456 for monitor_key in monitor_list {
457 res.push(self.read_channel_monitor_with_updates(
466 /// Read a single channel monitor, along with any stored updates for it.
468 /// It is extremely important that your [`KVStore::read`] implementation uses the
469 /// [`io::ErrorKind::NotFound`] variant correctly. For more information, please see the
470 /// documentation for [`MonitorUpdatingPersister`].
472 /// For `monitor_key`, channel storage keys be the channel's transaction ID and index, or
473 /// [`OutPoint`], with an underscore `_` between them. For example, given:
475 /// - Transaction ID: `deadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeef`
478 /// The correct `monitor_key` would be:
479 /// `deadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeef_1`
481 /// Loading a large number of monitors will be faster if done in parallel. You can use this
482 /// function to accomplish this. Take care to limit the number of parallel readers.
483 pub fn read_channel_monitor_with_updates<B: Deref, F: Deref>(
484 &self, broadcaster: &B, fee_estimator: &F, monitor_key: String,
485 ) -> Result<(BlockHash, ChannelMonitor<<SP::Target as SignerProvider>::EcdsaSigner>), io::Error>
487 B::Target: BroadcasterInterface,
488 F::Target: FeeEstimator,
490 let monitor_name = MonitorName::new(monitor_key)?;
491 let (block_hash, monitor) = self.read_monitor(&monitor_name)?;
492 let mut current_update_id = monitor.get_latest_update_id();
494 current_update_id = match current_update_id.checked_add(1) {
495 Some(next_update_id) => next_update_id,
498 let update_name = UpdateName::from(current_update_id);
499 let update = match self.read_monitor_update(&monitor_name, &update_name) {
500 Ok(update) => update,
501 Err(err) if err.kind() == io::ErrorKind::NotFound => {
502 // We can't find any more updates, so we are done.
505 Err(err) => return Err(err),
508 monitor.update_monitor(&update, broadcaster, fee_estimator, &self.logger)
512 "Monitor update failed. monitor: {} update: {} reason: {:?}",
513 monitor_name.as_str(),
514 update_name.as_str(),
517 io::Error::new(io::ErrorKind::Other, "Monitor update failed")
520 Ok((block_hash, monitor))
523 /// Read a channel monitor.
525 &self, monitor_name: &MonitorName,
526 ) -> Result<(BlockHash, ChannelMonitor<<SP::Target as SignerProvider>::EcdsaSigner>), io::Error> {
527 let outpoint: OutPoint = monitor_name.try_into()?;
528 let mut monitor_cursor = io::Cursor::new(self.kv_store.read(
529 CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE,
530 CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE,
531 monitor_name.as_str(),
533 // Discard the sentinel bytes if found.
534 if monitor_cursor.get_ref().starts_with(MONITOR_UPDATING_PERSISTER_PREPEND_SENTINEL) {
535 monitor_cursor.set_position(MONITOR_UPDATING_PERSISTER_PREPEND_SENTINEL.len() as u64);
537 match <(BlockHash, ChannelMonitor<<SP::Target as SignerProvider>::EcdsaSigner>)>::read(
539 (&*self.entropy_source, &*self.signer_provider),
541 Ok((blockhash, channel_monitor)) => {
542 if channel_monitor.get_funding_txo().0.txid != outpoint.txid
543 || channel_monitor.get_funding_txo().0.index != outpoint.index
547 "ChannelMonitor {} was stored under the wrong key!",
548 monitor_name.as_str()
551 io::ErrorKind::InvalidData,
552 "ChannelMonitor was stored under the wrong key",
555 Ok((blockhash, channel_monitor))
561 "Failed to read ChannelMonitor {}, reason: {}",
562 monitor_name.as_str(),
565 Err(io::Error::new(io::ErrorKind::InvalidData, "Failed to read ChannelMonitor"))
570 /// Read a channel monitor update.
571 fn read_monitor_update(
572 &self, monitor_name: &MonitorName, update_name: &UpdateName,
573 ) -> Result<ChannelMonitorUpdate, io::Error> {
574 let update_bytes = self.kv_store.read(
575 CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE,
576 monitor_name.as_str(),
577 update_name.as_str(),
579 ChannelMonitorUpdate::read(&mut io::Cursor::new(update_bytes)).map_err(|e| {
582 "Failed to read ChannelMonitorUpdate {}/{}/{}, reason: {}",
583 CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE,
584 monitor_name.as_str(),
585 update_name.as_str(),
588 io::Error::new(io::ErrorKind::InvalidData, "Failed to read ChannelMonitorUpdate")
592 /// Cleans up stale updates for all monitors.
594 /// This function works by first listing all monitors, and then for each of them, listing all
595 /// updates. The updates that have an `update_id` less than or equal to than the stored monitor
596 /// are deleted. The deletion can either be lazy or non-lazy based on the `lazy` flag; this will
597 /// be passed to [`KVStore::remove`].
598 pub fn cleanup_stale_updates(&self, lazy: bool) -> Result<(), io::Error> {
599 let monitor_keys = self.kv_store.list(
600 CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE,
601 CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE,
603 for monitor_key in monitor_keys {
604 let monitor_name = MonitorName::new(monitor_key)?;
605 let (_, current_monitor) = self.read_monitor(&monitor_name)?;
608 .list(CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE, monitor_name.as_str())?;
609 for update in updates {
610 let update_name = UpdateName::new(update)?;
611 // if the update_id is lower than the stored monitor, delete
612 if update_name.0 <= current_monitor.get_latest_update_id() {
613 self.kv_store.remove(
614 CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE,
615 monitor_name.as_str(),
616 update_name.as_str(),
626 impl<ChannelSigner: WriteableEcdsaChannelSigner, K: Deref, L: Deref, ES: Deref, SP: Deref>
627 Persist<ChannelSigner> for MonitorUpdatingPersister<K, L, ES, SP>
631 ES::Target: EntropySource + Sized,
632 SP::Target: SignerProvider + Sized,
634 /// Persists a new channel. This means writing the entire monitor to the
635 /// parametrized [`KVStore`].
636 fn persist_new_channel(
637 &self, funding_txo: OutPoint, monitor: &ChannelMonitor<ChannelSigner>,
638 _monitor_update_call_id: MonitorUpdateId,
639 ) -> chain::ChannelMonitorUpdateStatus {
640 // Determine the proper key for this monitor
641 let monitor_name = MonitorName::from(funding_txo);
642 // Serialize and write the new monitor
643 let mut monitor_bytes = Vec::with_capacity(
644 MONITOR_UPDATING_PERSISTER_PREPEND_SENTINEL.len() + monitor.serialized_length(),
646 monitor_bytes.extend_from_slice(MONITOR_UPDATING_PERSISTER_PREPEND_SENTINEL);
647 monitor.write(&mut monitor_bytes).unwrap();
648 match self.kv_store.write(
649 CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE,
650 CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE,
651 monitor_name.as_str(),
655 chain::ChannelMonitorUpdateStatus::Completed
660 "Failed to write ChannelMonitor {}/{}/{} reason: {}",
661 CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE,
662 CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE,
663 monitor_name.as_str(),
666 chain::ChannelMonitorUpdateStatus::UnrecoverableError
671 /// Persists a channel update, writing only the update to the parameterized [`KVStore`] if possible.
673 /// In some cases, this will forward to [`MonitorUpdatingPersister::persist_new_channel`]:
675 /// - No full monitor is found in [`KVStore`]
676 /// - The number of pending updates exceeds `maximum_pending_updates` as given to [`Self::new`]
677 /// - LDK commands re-persisting the entire monitor through this function, specifically when
678 /// `update` is `None`.
679 /// - The update is at [`CLOSED_CHANNEL_UPDATE_ID`]
680 fn update_persisted_channel(
681 &self, funding_txo: OutPoint, update: Option<&ChannelMonitorUpdate>,
682 monitor: &ChannelMonitor<ChannelSigner>, monitor_update_call_id: MonitorUpdateId,
683 ) -> chain::ChannelMonitorUpdateStatus {
684 // IMPORTANT: monitor_update_call_id: MonitorUpdateId is not to be confused with
685 // ChannelMonitorUpdate's update_id.
686 if let Some(update) = update {
687 if update.update_id != CLOSED_CHANNEL_UPDATE_ID
688 && update.update_id % self.maximum_pending_updates != 0
690 let monitor_name = MonitorName::from(funding_txo);
691 let update_name = UpdateName::from(update.update_id);
692 match self.kv_store.write(
693 CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE,
694 monitor_name.as_str(),
695 update_name.as_str(),
698 Ok(()) => chain::ChannelMonitorUpdateStatus::Completed,
702 "Failed to write ChannelMonitorUpdate {}/{}/{} reason: {}",
703 CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE,
704 monitor_name.as_str(),
705 update_name.as_str(),
708 chain::ChannelMonitorUpdateStatus::UnrecoverableError
712 let monitor_name = MonitorName::from(funding_txo);
713 // In case of channel-close monitor update, we need to read old monitor before persisting
714 // the new one in order to determine the cleanup range.
715 let maybe_old_monitor = match monitor.get_latest_update_id() {
716 CLOSED_CHANNEL_UPDATE_ID => self.read_monitor(&monitor_name).ok(),
720 // We could write this update, but it meets criteria of our design that calls for a full monitor write.
721 let monitor_update_status = self.persist_new_channel(funding_txo, monitor, monitor_update_call_id);
723 if let chain::ChannelMonitorUpdateStatus::Completed = monitor_update_status {
724 let cleanup_range = if monitor.get_latest_update_id() == CLOSED_CHANNEL_UPDATE_ID {
725 // If there is an error while reading old monitor, we skip clean up.
726 maybe_old_monitor.map(|(_, ref old_monitor)| {
727 let start = old_monitor.get_latest_update_id();
728 // We never persist an update with update_id = CLOSED_CHANNEL_UPDATE_ID
730 start.saturating_add(self.maximum_pending_updates),
731 CLOSED_CHANNEL_UPDATE_ID - 1,
736 let end = monitor.get_latest_update_id();
737 let start = end.saturating_sub(self.maximum_pending_updates);
741 if let Some((start, end)) = cleanup_range {
742 self.cleanup_in_range(monitor_name, start, end);
746 monitor_update_status
749 // There is no update given, so we must persist a new monitor.
750 self.persist_new_channel(funding_txo, monitor, monitor_update_call_id)
754 fn archive_persisted_channel(&self, funding_txo: OutPoint) {
755 let monitor_name = MonitorName::from(funding_txo);
756 let monitor = match self.read_monitor(&monitor_name) {
757 Ok((_block_hash, monitor)) => monitor,
760 match self.kv_store.write(
761 ARCHIVED_CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE,
762 ARCHIVED_CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE,
763 monitor_name.as_str(),
769 let _ = self.kv_store.remove(
770 CHANNEL_MONITOR_PERSISTENCE_PRIMARY_NAMESPACE,
771 CHANNEL_MONITOR_PERSISTENCE_SECONDARY_NAMESPACE,
772 monitor_name.as_str(),
778 impl<K: Deref, L: Deref, ES: Deref, SP: Deref> MonitorUpdatingPersister<K, L, ES, SP>
780 ES::Target: EntropySource + Sized,
783 SP::Target: SignerProvider + Sized
785 // Cleans up monitor updates for given monitor in range `start..=end`.
786 fn cleanup_in_range(&self, monitor_name: MonitorName, start: u64, end: u64) {
787 for update_id in start..=end {
788 let update_name = UpdateName::from(update_id);
789 if let Err(e) = self.kv_store.remove(
790 CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE,
791 monitor_name.as_str(),
792 update_name.as_str(),
797 "Failed to clean up channel monitor updates for monitor {}, reason: {}",
798 monitor_name.as_str(),
806 /// A struct representing a name for a monitor.
808 struct MonitorName(String);
811 /// Constructs a [`MonitorName`], after verifying that an [`OutPoint`] can
812 /// be formed from the given `name`.
813 pub fn new(name: String) -> Result<Self, io::Error> {
814 MonitorName::do_try_into_outpoint(&name)?;
817 /// Convert this monitor name to a str.
818 pub fn as_str(&self) -> &str {
821 /// Attempt to form a valid [`OutPoint`] from a given name string.
822 fn do_try_into_outpoint(name: &str) -> Result<OutPoint, io::Error> {
823 let mut parts = name.splitn(2, '_');
824 let txid = if let Some(part) = parts.next() {
825 Txid::from_str(part).map_err(|_| {
826 io::Error::new(io::ErrorKind::InvalidData, "Invalid tx ID in stored key")
829 return Err(io::Error::new(
830 io::ErrorKind::InvalidData,
831 "Stored monitor key is not a splittable string",
834 let index = if let Some(part) = parts.next() {
835 part.parse().map_err(|_| {
836 io::Error::new(io::ErrorKind::InvalidData, "Invalid tx index in stored key")
839 return Err(io::Error::new(
840 io::ErrorKind::InvalidData,
841 "No tx index value found after underscore in stored key",
844 Ok(OutPoint { txid, index })
848 impl TryFrom<&MonitorName> for OutPoint {
849 type Error = io::Error;
851 fn try_from(value: &MonitorName) -> Result<Self, io::Error> {
852 MonitorName::do_try_into_outpoint(&value.0)
856 impl From<OutPoint> for MonitorName {
857 fn from(value: OutPoint) -> Self {
858 MonitorName(format!("{}_{}", value.txid.to_string(), value.index))
862 /// A struct representing a name for an update.
864 struct UpdateName(u64, String);
867 /// Constructs an [`UpdateName`], after verifying that an update sequence ID
868 /// can be derived from the given `name`.
869 pub fn new(name: String) -> Result<Self, io::Error> {
870 match name.parse::<u64>() {
871 Ok(u) => Ok(u.into()),
873 Err(io::Error::new(io::ErrorKind::InvalidData, "cannot parse u64 from update name"))
878 /// Convert this monitor update name to a &str
879 pub fn as_str(&self) -> &str {
884 impl From<u64> for UpdateName {
885 fn from(value: u64) -> Self {
886 Self(value, value.to_string())
893 use crate::chain::ChannelMonitorUpdateStatus;
894 use crate::events::{ClosureReason, MessageSendEventsProvider};
895 use crate::ln::functional_test_utils::*;
896 use crate::util::test_utils::{self, TestLogger, TestStore};
897 use crate::{check_added_monitors, check_closed_broadcast};
898 use crate::sync::Arc;
899 use crate::util::test_channel_signer::TestChannelSigner;
901 const EXPECTED_UPDATES_PER_PAYMENT: u64 = 5;
904 fn converts_u64_to_update_name() {
905 assert_eq!(UpdateName::from(0).as_str(), "0");
906 assert_eq!(UpdateName::from(21).as_str(), "21");
907 assert_eq!(UpdateName::from(u64::MAX).as_str(), "18446744073709551615");
911 fn bad_update_name_fails() {
912 assert!(UpdateName::new("deadbeef".to_string()).is_err());
913 assert!(UpdateName::new("-1".to_string()).is_err());
917 fn monitor_from_outpoint_works() {
918 let monitor_name1 = MonitorName::from(OutPoint {
919 txid: Txid::from_str("deadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeef").unwrap(),
922 assert_eq!(monitor_name1.as_str(), "deadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeef_1");
924 let monitor_name2 = MonitorName::from(OutPoint {
925 txid: Txid::from_str("f33dbeeff33dbeeff33dbeeff33dbeeff33dbeeff33dbeeff33dbeeff33dbeef").unwrap(),
928 assert_eq!(monitor_name2.as_str(), "f33dbeeff33dbeeff33dbeeff33dbeeff33dbeeff33dbeeff33dbeeff33dbeef_65535");
932 fn bad_monitor_string_fails() {
933 assert!(MonitorName::new("deadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeef".to_string()).is_err());
934 assert!(MonitorName::new("deadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeef_65536".to_string()).is_err());
935 assert!(MonitorName::new("deadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeefdeadbeef_21".to_string()).is_err());
938 // Exercise the `MonitorUpdatingPersister` with real channels and payments.
940 fn persister_with_real_monitors() {
941 // This value is used later to limit how many iterations we perform.
942 let persister_0_max_pending_updates = 7;
943 // Intentionally set this to a smaller value to test a different alignment.
944 let persister_1_max_pending_updates = 3;
945 let chanmon_cfgs = create_chanmon_cfgs(4);
946 let persister_0 = MonitorUpdatingPersister {
947 kv_store: &TestStore::new(false),
948 logger: &TestLogger::new(),
949 maximum_pending_updates: persister_0_max_pending_updates,
950 entropy_source: &chanmon_cfgs[0].keys_manager,
951 signer_provider: &chanmon_cfgs[0].keys_manager,
953 let persister_1 = MonitorUpdatingPersister {
954 kv_store: &TestStore::new(false),
955 logger: &TestLogger::new(),
956 maximum_pending_updates: persister_1_max_pending_updates,
957 entropy_source: &chanmon_cfgs[1].keys_manager,
958 signer_provider: &chanmon_cfgs[1].keys_manager,
960 let mut node_cfgs = create_node_cfgs(2, &chanmon_cfgs);
961 let chain_mon_0 = test_utils::TestChainMonitor::new(
962 Some(&chanmon_cfgs[0].chain_source),
963 &chanmon_cfgs[0].tx_broadcaster,
964 &chanmon_cfgs[0].logger,
965 &chanmon_cfgs[0].fee_estimator,
967 &chanmon_cfgs[0].keys_manager,
969 let chain_mon_1 = test_utils::TestChainMonitor::new(
970 Some(&chanmon_cfgs[1].chain_source),
971 &chanmon_cfgs[1].tx_broadcaster,
972 &chanmon_cfgs[1].logger,
973 &chanmon_cfgs[1].fee_estimator,
975 &chanmon_cfgs[1].keys_manager,
977 node_cfgs[0].chain_monitor = chain_mon_0;
978 node_cfgs[1].chain_monitor = chain_mon_1;
979 let node_chanmgrs = create_node_chanmgrs(2, &node_cfgs, &[None, None]);
980 let nodes = create_network(2, &node_cfgs, &node_chanmgrs);
981 let broadcaster_0 = &chanmon_cfgs[2].tx_broadcaster;
982 let broadcaster_1 = &chanmon_cfgs[3].tx_broadcaster;
984 // Check that the persisted channel data is empty before any channels are
986 let mut persisted_chan_data_0 = persister_0.read_all_channel_monitors_with_updates(
987 &broadcaster_0, &&chanmon_cfgs[0].fee_estimator).unwrap();
988 assert_eq!(persisted_chan_data_0.len(), 0);
989 let mut persisted_chan_data_1 = persister_1.read_all_channel_monitors_with_updates(
990 &broadcaster_1, &&chanmon_cfgs[1].fee_estimator).unwrap();
991 assert_eq!(persisted_chan_data_1.len(), 0);
993 // Helper to make sure the channel is on the expected update ID.
994 macro_rules! check_persisted_data {
995 ($expected_update_id: expr) => {
996 persisted_chan_data_0 = persister_0.read_all_channel_monitors_with_updates(
997 &broadcaster_0, &&chanmon_cfgs[0].fee_estimator).unwrap();
998 // check that we stored only one monitor
999 assert_eq!(persisted_chan_data_0.len(), 1);
1000 for (_, mon) in persisted_chan_data_0.iter() {
1001 // check that when we read it, we got the right update id
1002 assert_eq!(mon.get_latest_update_id(), $expected_update_id);
1004 // if the CM is at consolidation threshold, ensure no updates are stored.
1005 let monitor_name = MonitorName::from(mon.get_funding_txo().0);
1006 if mon.get_latest_update_id() % persister_0_max_pending_updates == 0
1007 || mon.get_latest_update_id() == CLOSED_CHANNEL_UPDATE_ID {
1009 persister_0.kv_store.list(CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE,
1010 monitor_name.as_str()).unwrap().len(),
1012 "updates stored when they shouldn't be in persister 0"
1016 persisted_chan_data_1 = persister_1.read_all_channel_monitors_with_updates(
1017 &broadcaster_1, &&chanmon_cfgs[1].fee_estimator).unwrap();
1018 assert_eq!(persisted_chan_data_1.len(), 1);
1019 for (_, mon) in persisted_chan_data_1.iter() {
1020 assert_eq!(mon.get_latest_update_id(), $expected_update_id);
1021 let monitor_name = MonitorName::from(mon.get_funding_txo().0);
1022 // if the CM is at consolidation threshold, ensure no updates are stored.
1023 if mon.get_latest_update_id() % persister_1_max_pending_updates == 0
1024 || mon.get_latest_update_id() == CLOSED_CHANNEL_UPDATE_ID {
1026 persister_1.kv_store.list(CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE,
1027 monitor_name.as_str()).unwrap().len(),
1029 "updates stored when they shouldn't be in persister 1"
1036 // Create some initial channel and check that a channel was persisted.
1037 let _ = create_announced_chan_between_nodes(&nodes, 0, 1);
1038 check_persisted_data!(0);
1040 // Send a few payments and make sure the monitors are updated to the latest.
1041 send_payment(&nodes[0], &vec![&nodes[1]][..], 8_000_000);
1042 check_persisted_data!(EXPECTED_UPDATES_PER_PAYMENT);
1043 send_payment(&nodes[1], &vec![&nodes[0]][..], 4_000_000);
1044 check_persisted_data!(2 * EXPECTED_UPDATES_PER_PAYMENT);
1046 // Send a few more payments to try all the alignments of max pending updates with
1047 // updates for a payment sent and received.
1049 for i in 3..=persister_0_max_pending_updates * 2 {
1058 send_payment(&nodes[sender], &vec![&nodes[receiver]][..], 21_000);
1059 check_persisted_data!(i * EXPECTED_UPDATES_PER_PAYMENT);
1062 // Force close because cooperative close doesn't result in any persisted
1064 nodes[0].node.force_close_broadcasting_latest_txn(&nodes[0].node.list_channels()[0].channel_id, &nodes[1].node.get_our_node_id()).unwrap();
1066 check_closed_event(&nodes[0], 1, ClosureReason::HolderForceClosed, false, &[nodes[1].node.get_our_node_id()], 100000);
1067 check_closed_broadcast!(nodes[0], true);
1068 check_added_monitors!(nodes[0], 1);
1070 let node_txn = nodes[0].tx_broadcaster.txn_broadcast();
1071 assert_eq!(node_txn.len(), 1);
1073 connect_block(&nodes[1], &create_dummy_block(nodes[0].best_block_hash(), 42, vec![node_txn[0].clone(), node_txn[0].clone()]));
1075 check_closed_broadcast!(nodes[1], true);
1076 check_closed_event(&nodes[1], 1, ClosureReason::CommitmentTxConfirmed, false, &[nodes[0].node.get_our_node_id()], 100000);
1077 check_added_monitors!(nodes[1], 1);
1079 // Make sure everything is persisted as expected after close.
1080 check_persisted_data!(CLOSED_CHANNEL_UPDATE_ID);
1082 // Make sure the expected number of stale updates is present.
1083 let persisted_chan_data = persister_0.read_all_channel_monitors_with_updates(&broadcaster_0, &&chanmon_cfgs[0].fee_estimator).unwrap();
1084 let (_, monitor) = &persisted_chan_data[0];
1085 let monitor_name = MonitorName::from(monitor.get_funding_txo().0);
1086 // The channel should have 0 updates, as it wrote a full monitor and consolidated.
1087 assert_eq!(persister_0.kv_store.list(CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE, monitor_name.as_str()).unwrap().len(), 0);
1088 assert_eq!(persister_1.kv_store.list(CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE, monitor_name.as_str()).unwrap().len(), 0);
1091 // Test that if the `MonitorUpdatingPersister`'s can't actually write, trying to persist a
1092 // monitor or update with it results in the persister returning an UnrecoverableError status.
1094 fn unrecoverable_error_on_write_failure() {
1095 // Set up a dummy channel and force close. This will produce a monitor
1096 // that we can then use to test persistence.
1097 let chanmon_cfgs = create_chanmon_cfgs(2);
1098 let node_cfgs = create_node_cfgs(2, &chanmon_cfgs);
1099 let node_chanmgrs = create_node_chanmgrs(2, &node_cfgs, &[None, None]);
1100 let nodes = create_network(2, &node_cfgs, &node_chanmgrs);
1101 let chan = create_announced_chan_between_nodes(&nodes, 0, 1);
1102 nodes[1].node.force_close_broadcasting_latest_txn(&chan.2, &nodes[0].node.get_our_node_id()).unwrap();
1103 check_closed_event(&nodes[1], 1, ClosureReason::HolderForceClosed, false, &[nodes[0].node.get_our_node_id()], 100000);
1105 let mut added_monitors = nodes[1].chain_monitor.added_monitors.lock().unwrap();
1106 let update_map = nodes[1].chain_monitor.latest_monitor_update_id.lock().unwrap();
1107 let update_id = update_map.get(&added_monitors[0].1.channel_id()).unwrap();
1108 let cmu_map = nodes[1].chain_monitor.monitor_updates.lock().unwrap();
1109 let cmu = &cmu_map.get(&added_monitors[0].1.channel_id()).unwrap()[0];
1110 let test_txo = OutPoint { txid: Txid::from_str("8984484a580b825b9972d7adb15050b3ab624ccd731946b3eeddb92f4e7ef6be").unwrap(), index: 0 };
1112 let ro_persister = MonitorUpdatingPersister {
1113 kv_store: &TestStore::new(true),
1114 logger: &TestLogger::new(),
1115 maximum_pending_updates: 11,
1116 entropy_source: node_cfgs[0].keys_manager,
1117 signer_provider: node_cfgs[0].keys_manager,
1119 match ro_persister.persist_new_channel(test_txo, &added_monitors[0].1, update_id.2) {
1120 ChannelMonitorUpdateStatus::UnrecoverableError => {
1123 ChannelMonitorUpdateStatus::Completed => {
1124 panic!("Completed persisting new channel when shouldn't have")
1126 ChannelMonitorUpdateStatus::InProgress => {
1127 panic!("Returned InProgress when shouldn't have")
1130 match ro_persister.update_persisted_channel(test_txo, Some(cmu), &added_monitors[0].1, update_id.2) {
1131 ChannelMonitorUpdateStatus::UnrecoverableError => {
1134 ChannelMonitorUpdateStatus::Completed => {
1135 panic!("Completed persisting new channel when shouldn't have")
1137 ChannelMonitorUpdateStatus::InProgress => {
1138 panic!("Returned InProgress when shouldn't have")
1141 added_monitors.clear();
1143 nodes[1].node.get_and_clear_pending_msg_events();
1146 // Confirm that the `clean_stale_updates` function finds and deletes stale updates.
1148 fn clean_stale_updates_works() {
1149 let test_max_pending_updates = 7;
1150 let chanmon_cfgs = create_chanmon_cfgs(3);
1151 let persister_0 = MonitorUpdatingPersister {
1152 kv_store: &TestStore::new(false),
1153 logger: &TestLogger::new(),
1154 maximum_pending_updates: test_max_pending_updates,
1155 entropy_source: &chanmon_cfgs[0].keys_manager,
1156 signer_provider: &chanmon_cfgs[0].keys_manager,
1158 let persister_1 = MonitorUpdatingPersister {
1159 kv_store: &TestStore::new(false),
1160 logger: &TestLogger::new(),
1161 maximum_pending_updates: test_max_pending_updates,
1162 entropy_source: &chanmon_cfgs[1].keys_manager,
1163 signer_provider: &chanmon_cfgs[1].keys_manager,
1165 let mut node_cfgs = create_node_cfgs(2, &chanmon_cfgs);
1166 let chain_mon_0 = test_utils::TestChainMonitor::new(
1167 Some(&chanmon_cfgs[0].chain_source),
1168 &chanmon_cfgs[0].tx_broadcaster,
1169 &chanmon_cfgs[0].logger,
1170 &chanmon_cfgs[0].fee_estimator,
1172 &chanmon_cfgs[0].keys_manager,
1174 let chain_mon_1 = test_utils::TestChainMonitor::new(
1175 Some(&chanmon_cfgs[1].chain_source),
1176 &chanmon_cfgs[1].tx_broadcaster,
1177 &chanmon_cfgs[1].logger,
1178 &chanmon_cfgs[1].fee_estimator,
1180 &chanmon_cfgs[1].keys_manager,
1182 node_cfgs[0].chain_monitor = chain_mon_0;
1183 node_cfgs[1].chain_monitor = chain_mon_1;
1184 let node_chanmgrs = create_node_chanmgrs(2, &node_cfgs, &[None, None]);
1185 let nodes = create_network(2, &node_cfgs, &node_chanmgrs);
1187 let broadcaster_0 = &chanmon_cfgs[2].tx_broadcaster;
1189 // Check that the persisted channel data is empty before any channels are
1191 let persisted_chan_data = persister_0.read_all_channel_monitors_with_updates(&broadcaster_0, &&chanmon_cfgs[0].fee_estimator).unwrap();
1192 assert_eq!(persisted_chan_data.len(), 0);
1194 // Create some initial channel
1195 let _ = create_announced_chan_between_nodes(&nodes, 0, 1);
1197 // Send a few payments to advance the updates a bit
1198 send_payment(&nodes[0], &vec![&nodes[1]][..], 8_000_000);
1199 send_payment(&nodes[1], &vec![&nodes[0]][..], 4_000_000);
1201 // Get the monitor and make a fake stale update at update_id=1 (lowest height of an update possible)
1202 let persisted_chan_data = persister_0.read_all_channel_monitors_with_updates(&broadcaster_0, &&chanmon_cfgs[0].fee_estimator).unwrap();
1203 let (_, monitor) = &persisted_chan_data[0];
1204 let monitor_name = MonitorName::from(monitor.get_funding_txo().0);
1207 .write(CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE, monitor_name.as_str(), UpdateName::from(1).as_str(), &[0u8; 1])
1210 // Do the stale update cleanup
1211 persister_0.cleanup_stale_updates(false).unwrap();
1213 // Confirm the stale update is unreadable/gone
1216 .read(CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE, monitor_name.as_str(), UpdateName::from(1).as_str())
1220 nodes[0].node.force_close_broadcasting_latest_txn(&nodes[0].node.list_channels()[0].channel_id, &nodes[1].node.get_our_node_id()).unwrap();
1221 check_closed_event(&nodes[0], 1, ClosureReason::HolderForceClosed, false, &[nodes[1].node.get_our_node_id()], 100000);
1222 check_closed_broadcast!(nodes[0], true);
1223 check_added_monitors!(nodes[0], 1);
1225 // Write an update near u64::MAX
1228 .write(CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE, monitor_name.as_str(), UpdateName::from(u64::MAX - 1).as_str(), &[0u8; 1])
1231 // Do the stale update cleanup
1232 persister_0.cleanup_stale_updates(false).unwrap();
1234 // Confirm the stale update is unreadable/gone
1237 .read(CHANNEL_MONITOR_UPDATE_PERSISTENCE_PRIMARY_NAMESPACE, monitor_name.as_str(), UpdateName::from(u64::MAX - 1).as_str())
1241 fn persist_fn<P: Deref, ChannelSigner: WriteableEcdsaChannelSigner>(_persist: P) -> bool where P::Target: Persist<ChannelSigner> {
1246 fn kvstore_trait_object_usage() {
1247 let store: Arc<dyn KVStore + Send + Sync> = Arc::new(TestStore::new(false));
1248 assert!(persist_fn::<_, TestChannelSigner>(store.clone()));