Add tests for new deDupedAnnouncements struct in gossiper_test.
Test the various functionalities of the struct - that empty
struct contains no announcements, that announcements of each type
can be added and properly de-duplicated, that the batch of
announcements is delivered correctly, and that after reset the
struct again contains no announcements.
Add option to set trickleDelay for AuthenticatedGossiper in
command line, with default value of 300 milliseconds. Pass this
value to newServer, which uses it when creating a new instance of
AuthenticatedGossiper. Also set this value to 300 milliseconds when
creating nodes in integration tests.
For Part 1 of Issue #275. Create isolated private struct in
networkHandler goroutine that will de-duplicate
announcements added to the batch. The struct contains maps
for each of channel announcements, channel updates, and
node announcements to keep track of unique announcements.
The struct has a Reset method to reset stored announcements, an
AddMsg(lnwire.Message) method to add a new message to the current
batch, and a Batch method to return the set of de-duplicated
announcements.
Also fix a few minor typos.
In this commit, we fix an existing bug that could result in a panic if
we received a ChannelUpdate message with an unknown set of flags. If
the flag wasn’t set to zero or one, then the pubKey parameter would be
still nil when we attempted to validate it, causing an error to occur.
We remedy this by instead returning an error if the flags are unknown.
In a future commit, we will properly handle the set of flags that
indicates the channel should be disabled.
In this commit, we add a TCP fallback option for the
DNSSeedBootstrapper. We’ve received many reports of users unable to
bootstrap properly to the network due to the size of the SRV records we
currently return. It has been observed that many revolvers will simply
truncate and ignore the response due to the (current size).
To resolve (no pun intended) we now attempt to detect this failure mode
and will fallback to a manual TCP resolution in the case that our SRV
query over UDP fails. We do this by querying the special record at the
"soa." sub-domain of supporting DNS servers. The retuned IP address
will be the IP address of the authoritative DNS server. Once we have
this IP address, we'll connect manually over TCP to request the SRV
record. This is necessary as the records we return are currently too
large for a class of resolvers, causing them to be filtered out.
This commit refactors the SynchronizeNode logic such that
it can be called without interacting with the gossiper's
main execution loop. This method does not require access
to any of the gossiper's internal state, making the change
fairly straightforward. The primary motivation behind
this change is to minimize the possibility of introducing
deadlock scenarios between the gossiper and server.
This commit fixes a slight bug in the announcement processing logic
within the AuthenticatedGossiper. Before this commit, it was possible
for us to ignore one of our now announcements due to it being
pre-mature, rendering (atm) the channel unusable by the ChannelRouter
itself. To fix this, we know only check for a pre mature announcement
iff the message is coming from a remote node.
This commit adds an additional return value to the updateChannel
method. We also now return the original ChannelAnnouncement, this can
be useful as it let’s the caller ensure that the channel announcement
will be broadcast along side the new channel update.
This commit does two things. First we fix a deadlock bug within boltdb
that could arise when the channel router attempted to open a
transaction, while the retransmission loop was attempting to send a
message to the channel router (circular wait). Second, we’ve refactored
out the retransmission into its own function. This allows us to kick
off retransmission as soon as the AuthenticatedGossiper is created.
This commit modifies the recently modified logic for self-channel
retransmission to exclude pruning *our* channels which haven’t been
updated since the broadcastInterval. Instead, we only re-broadcast
channels of ours that haven’t been updated in 24 hours.
This commit adds a new interface the to discovery package:
NetworkPeerBootstrapper. The NetworkPeerBootstrapper interface is meant
to be used to bootstrap a new peer joining the network to the set of
existing active peers within the network. Callers are encouraged to
utilize several boostrappers in series as redundant sources of
information. The MultiSourceBootstrap function will takes a set of
boostrappers, and compose their outputs into a single unified set of
addresses.
Two concrete implementations of the NetworkPeerBootstrapper interface
have been added as a part of this commit: the ChannelGraphBootstrapper
and the DNSSeedBootstrapper. The former will utilize the authenticated
node advertisements within the calling nodes view to boostrap new
connections. The latter will use a set of BOLT-0010 compliant DNS seeds
to query. This DNS seeding more will likely be used by nodes initial
joining the network, as they may not yet have the channel graph as they
haven’t connected to any peers.
This commit fixes an existing bug in the announcement processing of the
authenticated gossiper. We will now only examine an incoming
announcement for validity if it targets the chain that we’re currently
active on.
This commit ensures that we *always* populate the ChainHash field when
we’re crafting new channel update messages either due to the periodic
broadcast, or when we’re updating the routing policies of a set of
target channels. Previously, this wasn’t set which would cause nodes to
reject the newly crafted ChannelUpdate messages.
This commit adds a new method to the AuthenticatedGossiper:
PropagateFeeUpdate. This new method will allow callers to update the
fee schedule advertised for a particular channel, or all currently
active channels. With this method exposed, the AuthenticatedGossiper
will now craft the new channel update messages, sign the new state,
commit the new policy to the underlying graph, and finally add the
message to the next announcement epoch batch.
This commit implements the simplification within the latest version of
BOLT-0007. With this change, we simply sign the message hash directly
with the bitcoin keys, rather than signing the node keys themselves.
According to BOLT#7, nodes will ignore node announcements
for nodes not found in any previous channel announcements.
This commit makes the discovery service send its known
channels before its known nodes when syncing the channel
graph with a peer.
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.
In this commit waiting proofs array have been replaced with persistant
boltd storage which removes the possibility for the half proof to be
lost during half proof exchange.
In case of the situation when we receive remote announcement with
channel id which pointing out to unknown channel the announcement have
been silently rejected. Now such announcement will be added to the
waiting proof map.
Such solution has a serious drawback - by adding the announcement proof
without information about channel itself (that this announcement have
been received from the node eligible to sending it), we allow
overwriting the waiting proof map by `Eve` node.
This commit modifies the SignAnnouncement helper function to instead
take a MessageSigner interface rather than the deprecated MessageSigner
struct. We also modify the function to explicitly take the identity
public key to use when signing.
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.
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
