In this commit we modify the newRoute function to also add the source
node to the nextHopMap index. With this addition the indexes will now
allow the router to react based on failures that occur during the
_first_ hop, meaning the channel directly attached to the source node.
This commit adds three new indexes to the Route struct. These indexes
allow a caller to check if a channel is in the route, check if a node
is in the route, query the next node after a target node, and query the
next channel after a target node. The combination of these new indexes
will allow the ChannelRouter to prune away routes from the available
set in response to any received errors.
This commit implements 2-week zombie channel pruning. This means that
every GraphPruneInterval (currently set to one hour), we’ll scan the
channel graph, marking any channels which haven’t had *both* edges
updated in 2 weeks as a “zombie”. During the second pass, all “zombie”
channel are removed from the channel graph all together.
Adding this functionality means we’ll ensure that we maintain a
“healthy” network view, which will cut down on the number of failed
HTLC routing attempts, and also reflect an active portion of the graph.
Use binary.Read/Write in functions to serialize and deserialize
channel close summary and HTLC boolean data, as well as in
methods to put and fetch channel funding info. Remove lnd
implementations of readBool and writeBool as they are no
longer needed. Also fix a few minor typos.
Use sort.Slice in FindRoutes function in routing/router.go, as part
of the move to use new language features. Remove sortableRoutes type
wrapper for slice of Routes since it is no longer needed to sort routes.
In this commit we modify the existing
TestSendPaymentRouteFailureFallback to use a non-critical error aside
from FailChannelDisabled. This is necessary as the behavior of the
current error handling can fail due to us sending in a nil error.
This commit modifies the way we currently interpret errors when sending
payments via the SendToSwitch method. We split the errors into two
broad sections: critical errors which cause us to abandon the payment
dispatch all together, and errors which are transient meaning we should
continue trying to remainder of the returned routes.
Note that we haven’t yet properly implemented all the necessary
measures such as filtering edges that are detected as being temporarily
inactive, etc.
This change should correct erroneous behavior such as continuing to try
all available routes in the face of an invalid payment hash error and
the like.
This commit modifies the way we do path caching. Rather than only
caching within SendPayment, we now cache routes within FindRoutes. This
is more natural as SendPayment eventually calls FindRoute. As a result
of this commit, queries to FindRoute are now properly cached, speeding
up applications which are focused on graph visualization or querying
rather than sending payments.
This commit reduces the neutrino.WaitForMoreCFHeaders parameter when
instantiating a neutrino instance as a lower value will allow the tests
to complete more quickly.
This commit fixes an oversight in the path finding code when converting
a path into a route. Currently, for the last hop, we’d emplace the
expiry delta of the last hop within the per-hop payload. This was left
over from a prior version of the specification.
To fix this, we’ll now emplace the _absolute_ final HTLC expiry with
the payload, such that, the final hop that verify that the HTLC has not
been tampered with in flight.
This commit fixes an lingering bug within the path finding logic of the
router. Previously we used the edge policy directly attached to the
outgoing channel of the node we were traversing to calculate the fees
and time lock information. This is incorrect, as we instead should be
using the policy of the *connecting* node as we’ll need to pay for
transit as they dictate.
To remedy this, we now grab the incoming+outgoing edges and use those
accordingly when building the initial path.
This commit makes a precautionary change in order to ensure that the
upper bound on the number of iteration’s within our version of Yen’s
algorithm is fixed.
This commit makes the routing cache invalidation a bit more aggressive.
We now invalidate the cache on each new block as the routes in the
cache are based on the current block height. Using the cached items may
cause our routes to fail due to them having time locks which have
already expired.
This commit implements some missing functionality, namely before all
time locks were calculated off of a base height of 0 essentially.
That’s incorrect as all time locks within HTLC’s would then be already
expired. We remedy this requesting the latest height when creating a
route to ensure that our time locks are set properly.
This commit introduces the requirement specified in BOLT#7,
where we ignore any node announcements for a specific node
if we yet haven't seen any channel announcements where this
node takes part. This is to prevent someone DoS-ing the
network with cheap node announcements. In the router this
is enforced by requiring a call to AddNode(node_id) to
be preceded by an AddEdge(edge_id) call, where node_id is
one of the nodes in edge_id.
Modifies the test cases in `TestEdgeUpdateNotification` and
`TestNodeUpdateNotification` to check for the possibility of notifications
being delivered out of order. This addresses some sporadic failures that
were observed when running the test suite.
I looked through some of the open issues but didn't see any addressing this
issue in particular, but if someone could point me to any relevant issues
that would be much appreciated!
Issue
-----
Currently the test suite validates notifications received in the order they
are submitted. The check fails because the verification of each
notification is statically linked to the order in which they are delivered,
seen
[here](1be4d67ce4/routing/notifications_test.go (L403))
and
[here](1be4d67ce4/routing/notifications_test.go (L499))
in `routing/notifications_test.go`. The notifications are typically
delivered in this order, but causes the test to fail otherwise.
Proposed Changes
-------------------
Construct an index that maps a public key to its corresponding edges and/or
nodes. When a notification is received, use its identifying public key and
the index to look up the edge/node to use for validation. Entries are
removed from the index after they are verified to ensure that the same
entry is validated twice. The logic to dynamically handle the verification
of incoming notifications rests can be found here
[here](https://github.com/cfromknecht/lnd/blob/order-invariant-ntfns/routing/notifications_test.go#L420)
and
[here](https://github.com/cfromknecht/lnd/blob/order-invariant-ntfns/routing/notifications_test.go#L539).
Encountered Errors
--------------------
* `TestEdgeUpdateNotification`: notifications_test.go:379: min HTLC of
edge doesn't match: expected 16.7401473 BTC, got 19.4852751 BTC
* `TestNodeUpdateNotification`: notifications_test.go:485: node identity
keys don't match: expected
027b139b2153ac5f3c83c2022e58b3219297d0fb3170739ee6391cddf2e06fe3e7, got
03921deafb61ee13d18e9d96c3ecd9e572e59c8dbd0bb922b5b6ac609d10fe4ee4
Recreating Failing Behavior
---------------------------
The failures can be somewhat difficult to recreate, I was able to reproduce
them by running the unit tests repeatedly until they showed up. I used the
following commands to bring them out of hiding:
```
./gotest.sh -i
go test -test.v ./routing && while [ $? -eq 0 ]; do go test -test.v ./routing; done
```
I was unable to recreate these errors, or any others in this package, after
making the proposed changes and leaving the script running continuously for
~30 minutes. Previously, I could consistently generate an error after ~20
seconds had elapsed on the latest commit in master at the time of writing:
78f6caf5d2e570fea0e5c05cc440cb7395a99c1d. Moar stability ftw!
Within the network, it's important that when an HTLC forwarding failure
occurs, the recipient is notified in a timely manner in order to ensure
that errors are graceful and not unknown. For that reason with
accordance to BOLT №4 onion failure obfuscation have been added.
The btclog package has been changed to defining its own logging
interface (rather than seelog's) and provides a default implementation
for callers to use.
There are two primary advantages to the new logger implementation.
First, all log messages are created before the call returns. Compared
to seelog, this prevents data races when mutable variables are logged.
Second, the new logger does not implement any kind of artifical rate
limiting (what seelog refers to as "adaptive logging"). Log messages
are outputted as soon as possible and the application will appear to
perform much better when watching standard output.
Because log rotation is not a feature of the btclog logging
implementation, it is handled by the main package by importing a file
rotation package that provides an io.Reader interface for creating
output to a rotating file output. The rotator has been configured
with the same defaults that btcd previously used in the seelog config
(10MB file limits with maximum of 3 rolls) but now compresses newly
created roll files. Due to the high compressibility of log text, the
compressed files typically reduce to around 15-30% of the original
10MB file.
This commit fixes a send on closed channel panic by adding additional
synchronization when cancelling the notifications for a particular
topology client. We now ensure that all goroutines belonging to a
particular topology client exit fully before we close the notification
channel in order to avoid a panic.
This commit adds a new method to the routing.Route struct:
ToHopPayloads. This function will converts a complete route into the
series of per-hop payloads that is to be encoded within each HTLC using
an opaque Sphinx packet.
We can now use this function when creating the sphinx packet to
properly encoded the hop payload for each hop in the route.
This commit inches towards fully validation+adherance of the per-hop
payloads within an HTLC’s route by properly calculating the outgoing
time lock value for each hop according to the current draft
specification.
This commit fixes a possible race condition wherein a call to
FilterBlock after a call to UpdateFilter would result in the call to
FilterBlock not yet using the updated filter. We fix this by ensuring
the internal chain filter is updated by the time the call to
FilterBlock returns.
This commit optimizes the neutrino implementation of FilterBlock method
of the ChainView interface. The old implementation would _always_ fetch
the entire block and manually scan through it. Instead, we can just
fetch the filter, and then if the items match, fetch the block itself.
This will save bandwidth during a lnd node’s pruning of the channel
graph after a period of dormancy.
This commit adds an initial rough implementation father ChainNotifier
interface for neutrino, our new light client implementation. This
implementation largely borrows from the existing BtcdNotifier
implementation. As a result, a follow up commit will perform two
refactoring in order to further consolidate code.
This commit adds a new implementation of the FilteredChainView
interface. This implementation speaks purely to the p2p network and is
backed by a new experimental light client implementation.
This commit replaces the hard-coded 5000 satoshi fees with calls to the
FeeEstimator interface. This should provide a way to cleanly plug in
additional fee calculation algorithms in the future. This change
affected quite a few tests. When possible, the tests were changed to
assert amounts sent rather than balances so that fees wouldn't need to
be taken into account. There were several tests for which this wasn't
possible, so calls to the static fee calculator were made.
This commit fixes a panic due to a send on a closed channel that could
possibly occur depending on the order of channel closes when a client
goes to cancel a topology notification client.
Previously we closed the ntfnChan first, this would possible result in
a panic as the goroutine may have succeeded on a send at the same time
the channel was closed. Instead, we now close the `exit` channel first
which is meant to be a signal to the goroutine that the client has been
canceled.
This commit modifies the processing in the routing package eo new
announcements. Previously, if we cgot a cnew channel announcement but
didn’t yet know of the verses that the chanell connected, the
cnnounacment would be accepted. This behavior was eronoues as if the
channel were to be queried for, the DB query would fail as we would be
unable to retrieve the two nodes involved int he channel.
To avoid such an error case, we will now _reject_ any channel
announcements in which we don’t yet have a valid node announcement for
the connected nodes. This case has been inserted into the handling of
channel announcement, a new test has been added, and finally older
tests have also been updated to ensure that nodes are added to the
database _before_ the edge is.
This commit modifies the routing package to no longer use the
ChainNotifier for pruning the channel graph. Instead, we now use the
FilteredChainView interface to more (from the ChannelRouter’s PoV)
efficiently maintain the channel graph.
Rather than scanning the _entire_ block manually, we now rely on the
FilteredChainView to provide us with FilteredBlocks which include
_only_ the relevant transactions that we care about.
This commit adds a new set of behavioral interface level tests to the
chain view package. This set of tests can now be used in order to check
proper conformity to this “specification” for all future
implementations of the chain view package.
This commit adds the first concrete implementation of the
chainview.FilteredChainView interface. The implementation of this
interface, BtcdFilteredChainView is backed by a web sockets connection
to an active btcd instance.
This commit creates a new package as sub-package within the routing
package: chainview. This package is centered around a single interface
definition: the FilteredChainView. This interface is to be used to
allow the routing package to watch a _subset_ of the UTXO set for any
modifications. In the case of LN, the subset of the UTXO set that we
care about is the set of currently opened channels.
In a future commit the routing package will be modified to remove the
current full block scanning with processing of FilteredBlock
notification, and proper updates to the filter as observed by the
FilteredChainView.
This commit fixes a pretty nasty unnoticed bug within the main
k-shortest paths algorithm loop. After a new candidate path is found,
the rootPath (the path up to the pivot node) and the spurPath (the
_new_ path after the pivot node) are to be combined into a new candiate
shortest path. The prior logic simply appended the spurPath onto the
end of the rootPath to create a slice. However, if the case that the
currnet rootPath is really a sub-path in a larger slice, then this will
mutate the underlying slice.
This bug would manifest when doing path finding and cause an infinite
loop as the slice kept growing with new spurPaths, causing the loop to
never terminate. We remedy this bug by properly create a new backing
slice, and adding the elements to them rather than incorrectly mutating
an underlying slice.
This commit fixes a bug within the k-shortest paths routine which could
result in a daemon panic when traversing a graph with particular
characteristics. Before referencing the path to create a sub-slice, we
we’re properly asserting that the length of the path was at least as
long as the current rootPath in question. We fix this by simply
ensuring the length of the slice is adequate before proceeding with the
operation.
This commit implements some minor coding style, commenting and naming
clean up after the recent major discovery service was merged into the
codebase.
Highlights of the naming changes:
* fundingManager.SendToDiscovery -> SendAnnouncement
* discovery.Discovery -> discovery.AuthenticatedGossiper
The rest of the changes consist primary of grammar fixes and proper
column wrapping.
Originally we adding the edge without proof in order to able to use it
for payments path constrcution. This method will allow us to populate
the announcement proof after the exchange of the half proofs and
constrcutrion of full proof finished.
Change the name of fields of messages which are belong to the discovery
subsystem in a such way so they were the same with the names that are
defined in the specification.
Add usage of the 'discovery' package in the lnd, now discovery service
will be handle all lnwire announcement messages and send them to the
remote party.
In this commit the routing package was divided on two separete one,
this was done because 'routing' package start take too much responsibily
on themself, so with following commit:
Routing pacakge:
Enitites:
* channeldb.ChannelEdge
* channeldb.ChannelPolicy
* channeldb.NodeLightning
Responsibilities:
* send topology notification
* find payment paths
* send payment
* apply topology changes to the graph
* prune graph
* validate that funding point exist and corresponds to given one
* to be the source of topology data
Discovery package:
Entities:
* lnwire.AnnounceSignature
* lnwire.ChannelAnnouncement
* lnwire.NodeAnnouncement
* lnwire.ChannelUpdateAnnouncement
Responsibilities:
* validate announcement signatures
* sync topology with newly connected peers
* handle the premature annoucement
* redirect topology changes to the router susbsystem
* broadcast announcement to the rest of the network
* exchange channel announcement proofs
Before that moment all that was in the 'routing' which is quite big for
one subsystem.
split
This commit modifies address handling in the NodeAnnouncement struct,
switching from net.TCPAddr to []net.Addr. This enables more flexible
address handling with multiple types and multiple addresses for each
node. This commit addresses the first part of issue #131 .
This commit fixes the issue of broken builds in versions other than go
1.7.5 by sorting according to the sort.Interface interface rather than
the newly available sort.Slice function.
This commit adds caching to our route finding. Caching is done on a
tuple-basis mapping a (dest, amt) pair to a previously calculated set
of shortest paths. The cache invalidated on two occasions: when a block
closes a set of transactions, or we received a new channel update or
channel announcement message.
With this change, payments are now snappier from the PoV of an
application developer as we no longer need to do a series of disk-seeks
before we dispatch each payment.
This commit adds payment route failure fallback to SendPayment. By
this, we mean that we now take all the possible routes found during
path finding and try them in series. Either a route fails and we move
onto the next one, or the route is successful and we terminate early.
With this commit, sending payments using lnd is now much more robust as
if there exists an eligible route with sufficient capacity, it will be
utilized.
This commit modifies the existing FindRoute method on the ChannelRouter
to now use the KSP implementation added in a prior commit.
This new method FindRoutes, is able to find all the possible paths
between a source and destination. The method takes all paths reported
by findPaths, and attempt to turn each of them into a route. A route
differs from a path in that is has complete time-lock and fee
information. Some paths may not be able to be turned into routes as
once fees are accounted for the have an insufficient flow. We then take
the routes, sort them by total fee (with time-lock being a
time-breaker), then return them in sorted order.
With this commit we make our routing more robust by looking for the
k-shortest paths rather than a single shortest path and using that
unconditionally. In a nut shell Yen’s algorithm does the following:
* Find the shortest path from the source to the destination
* From k=1…K, walk the kth shortest path and find possible
divergence from each node in the path
Our version of Yen’s implemented is slightly modified, rather than
actually increasing the edge weights or remove vertexes from the graph,
we instead use two black-lists: one for edges and the other for
vertexes. Our modified version of Djikstra’s algorithm is then made
aware of these black lists in order to more efficiently implement the
path iteration via spur and root node.
This commit modifies the findRoute method by first calling it findPath,
but also making the following modifications.
First, two new black-listing maps are now passed in. These two maps
contain vertexes but also edges to ignore while performing path
finding. These maps will be used in order to ensure that we don’t
duplicate paths or back-track when executing our KSP algorithm.
Next, we now ensure that the path returned from the findPath function
is ordered properly in the direction of source to target. Such a change
is required for our KSP algorithm to function properly.
This commit adds a new heap structure to heap.go which will be used for
storing candidate paths during the iterations of the k-shortest paths
algorithm.
This commit modifies the findRoute function to decouple the
validation+creation of a route, from the path finding algorithm itself.
When we say “route”, we mean the full payment route complete with
time-lock and fee information. When we say “path” we simple mean an
ordered set of channel edges from one node to another target node.
With this commit we can now perform path finding independent of route
creation which will be needed in the up coming refactor to implement a
new modified k-shortest paths algorithm.
This commit slightly modified findRoute to accept the node which should
be used as the starting point in our path finding algorithm. With this
change, as we move to a k-shortest paths algorithm this modification
will be needed as all of our path finding attempts won’t always
originate from a the same starting point.
In this commit we now utilize the node distance heap that was added in
a prior commit into our core path finding logic. With this new data
structure, we no longer linearly scan the distance of all vertexes from
the source node when deciding which one to greedily explore.
Instead, we now start with the source added to our distance heap, then
new vertexes are progressively added to our heap as their edges are
explored. With this change, we move the computational complexity of our
path finding algorithm closer to the theoretical limit.
This commit modifies our modified version of Dijkstra's to include
sufficient link capacity within the algorithm’s relaxation condition.
With this change, we’ll now avoid exploring the graph in the direction
of a link that has insufficient capacity, allowing us to terminate path
finding more quickly.
This commit introduces a new heap struct that will be used to keep
track of the next closest node to the source during path finding within
our modified Dijkstra's algorithm.
This commit fixes a bug which was originally introduced when the
topology notifications were added to the channel router. The issue was
that a pointer to the loop-scope range variable was being passed into
the goroutine which dispatches the notification rather than the value
itself. It seems that the memory location is re-used between range
iterations causing the same client to receive _all_ the notifications.
This bug is fixed by passing a copy of the client struct rather than a
pointer to the range variable.
In the process, we also add some additional debug logging messages, and
remove the Curve parameter from any public keys involved in a
notification so the pretty print properly.
This commit modifies the `ChannelEdgeUpdate` struct to include the
channel point itself within the notifications. Such a change improves
the notificaiton experience for callers as it allows them to filter out
update notifications based on a familiar object within the codebase: a
channel point.
