This commit ensures htlcs currently held by the interceptor
Are not sent again. This prevents potential races in the user code that
may lead to loosing funds.
In this commit the location of where chain control services
are stopped is shifted to be closer to the point they are started.
Stopping of two services: "wallet" and "feeEstimator" that are started
inside the "newChainControlFromConfig" was shifted from server.go to
the cleanup function.
In addition the chainView.Stop was also removed from the server.Stop as
it is already handled by the router, where it is being started.
In this commit every succesfull started sub system is tracked
and in case of a failure to complete the Start function all
these sub systems are stopped using a generic cleanup function.
In order to be consistent with other sub systems an error is now
returned from the Stop functions.
This also allows writing a generic cleanup mechanism to stop all
sub systems in case of a failure.
Adding persistency to neutrino re-scan progress. In case the re-scan
failed in the middle or the node was restarted before the re-scan
has been completed this fix will start from the last point instead
of the beginning.
On long re-scan , such as old channels on mobile devices this is
crucial.
In this commit we add the ability to intercept forwarded htlc packets
straight from the RPC layer. The RPC layer handles a bidrectional stream
that comminucates to the client the intercepted packets and handles its
response by coordinating with the interceptable switch.
In this commit we implement a wrapper arround the switch, called
InterceptableSwitch. This kind of wrapper behaves like a proxy which
intercepts forwarded packets and allows an external interceptor to
signal if it is interested to hold this forward and resolve it
manually later or let the switch execute its default behavior.
This infrastructure allows the RPC layer to expose interceptor
registration API to the user and by that enable the implementation
of custom routing behavior.
As part of the preparation to the switch interceptor feature, this
function is changed to return error instead of error channel that
is closed automatically.
Returning an error channel has become complex to maintain and
implement when adding more asynchronous flows to the switch.
The change doesn't affect the current behavior which logs the
errors as before.
This commit introduces an optimization for the chain scanning
used in historical confirmations notifications.
It now uses batch request instead of requesting filters one by one.
This commit enables the user to specify he is not interested in
automatically close channels with pending payments that their
corresponding htlcs have timed-out.
By requiring a configurable grace period uptime of our node
before closing such channels, we give a chance to the other node to
properly cancel the htlc and avoid unnecessary on-chain transaction.
In mobile it is very important for the user experience as otherwise
channels will be force closed more frequently.
This commit changes the shouldGoOnChain signature to get the htlc
as parameter. I will allow the function to take decisions based on
whether the htlc is Incoming or Outgoing.
Refresh channel memory state whenever the short channel id is refreshed.
This is to make the in-memory channel consistent with the disk data.
Fixes#3765.
The current approach iterates all channels in the graph in order to
filter those in need. This approach is time consuming, several seconds
on my mobile device for ~40,000 channels, while during this time the
db is locked in a transaction.
The proposed change is to use an existing functionality that utilize the
fact that channel update are saved indexed by date. This method enables
us to go over only a small subset of the channels, only those that
were updated before the "channel expiry" time and further filter
them for our need.
The same graph that took several seconds to prune was pruned, after
the change, in several milliseconds.
In addition for testing purposes I added Initiator field to the
testChannel structure to reflect the channeldEdgePolicy direction.
This commit adds an index bucket, disabledEdgePolicyBucket, for those
ChannelEdgePolicy with disabled bit on.
The main purpose is to be able to iterate over these fast when prune is
needed without the need for iterating the whole graph.
The entry points for accessing this index are:
1. When updating ChannelEdgePolicy - insert an entry.
2. When deleting ChannelEdge - delete the associated entries.
3. When querying for disabled channels - implemented DisabledChannelIDs
function
In this commit two fields were added to the Channl RPC result in both
open and pending states.
The fields: local_chan_reserve, remote_chan_reserve represents the
reservation the nodes are rquired to keep in both sides of the channel.
This is usefull when calculating the "real" inbound and outbound
liquidity in an accurate way.
This commit fixes a but where restarting LND at the same process causes
It to fail.
The problem resides in the fact that an array of permissions is
initialized as a package variable and when creating the
RPCServer all subserver permissions are appended while checking for
duplicates.
On subsequent restart this array is left over from the previous run and
being populated again with the same permissions causing a duplicate
error.
The solution is simple, just to extract out the initial permissions to
a separate function and call it from the context is is needed.
This commit modified the condition to whether drop an existing
connection to a peer when a new connection to this peer is
established.
The previous algorithm used public keys comparison for this decision
which determines that between every two nodes only one of them will
ever drop the connection in such cases.
The problematic case is when a node disconnects and reconnects in a
short interval which is the case of mobile devices.
In such case it takes as much as the "timeout" configured value for
the remote node to detect the "disconnection" (and try to reconnect
if this connection is persistent). In the case this node is also the
one that has the "smaller" public key the reconnect attempts of the
other node will be rejected causing it impossible to fast reconnect.
The solution is to only drop the connection if if we already have a
connected peer that is of the opposite direction from the this new
connection. By doing so the "initiator" will be enabled to replace
the connection and recconnect immediately.
Roei Erez