Previously, the Switch would not check waiting-close channels' fwdpkgs
for settles or fails to reforward. This could result in a force close
in a rare edge case if a restart occurred at the wrong time. Now,
waiting-close fwdpkgs are checked and the issue is avoided.
In order to reduce the number of calls to the db we try to process as
few channels as we can + try to not do extra work for each of them.
- First fetch all the channels. Then, filter all the public ones and
sort the potential candidates by remote balance.
- Filter out each potential candidate as soon as possible.
- Only check the alias if the channel supports scid aliases.
- Because we sort the channels by remote balance, we will hit the
target amount, if possible, as soon as we can.
We do not want to leak information about our remote balances, so we
shuffle the hop hints (the forced ones go always first) so the invoice
receiver does not know which channels have more balance than others.
It's possible that a user might not want the Tor private key to sit on the disk in plaintext (it is a private key after all). So this commit adds a new flag to encrypt the Tor private key on disk using the wallet's seed. When the --tor.encryptkey flag is used, LND will still write the Tor key to the same file, however it will now be encrypted intead of plaintext. This essentially uses the same method to encrypt the Tor private key as is used to encrypt the Static Channel Backup file.
The functions inside of the crypto.go file in chanbackup (like EncryptPayloadToWriter and DecryptPayloadFromReader) can be used by a lot of things outside of just the chanbackup package. We can't just reference them directly from the chanbackup package because it's likely that it would generate circular dependencies. Therefore we need to move these functions into their own package to be referenced by chanbackup and whatever new functionality that needs them
In this commit, we catch and error on the specific case where an
openchannel request has a zero-value local funding amount. This is just
to ensure that the error message returned makes more sense.
NewForce's Pause method doesn't reset the ticker, so a test flake
would occur in TestChannelLinkCancelFullCommitment where
PendingCommitTicker.Pause() was called, but the underlying timer was
still ticking. When PendingCommitTicker.Resume() was called, an
unlucky Ticks() call could end up firing, leading to the link being
shut down.
This fixes a bug where a caller would:
- call NotifyWhenOnline and pass in a channel
- the server's mutex is held and the pubStr is checked in peersByPub
- the peer object is in the peersByPub map
- the peer object has its quit channel closed
- QuitSignal is called in the select statement, and the function
returns
- the caller is still waiting for the channel to be sent on forever
since it was never sent on or added to the peerConnectedListeners
slice.
This patch fixes the above bug by adding the channel to the
peerConnectedListeners slice if the QuitSignal select case is called.
localDiscoverySignals needs to be guarded by its mutex. I was unable to
write a test case that would trigger a race under the race detector, but
better safe than sorry.
Lists with non-primitive members aren't supported in the query string of
a GET request with the current version of the grpc-gateway library. To
allow route_hints to be set through REST, we also offer a POST endpoint
for that call where the encoding of the request parameter can be
specified as JSON.
Olaoluwa Osuntokun