[anchors] cap value reserved for anchor fee bumping #5274
Conversation
|
Let's also return to default anchors in this PR? |
halseth
force-pushed
the
anchors-reserved-value-max
branch
2 times, most recently
from
e47c105
to
362b204
Compare
Done. |
| WumboChans bool `long:"wumbo-channels" description:"if set, then lnd will create and accept requests for channels larger chan 0.16 BTC"` | ||
|
|
||
| // Anchors enables anchor commitments. | ||
| // TODO(halseth): transition itests to anchors instead! |
There was a problem hiding this comment.
IIRC we had a branch somewhere the (started to?) did this?
There was a problem hiding this comment.
Yeah, I had a draft PR for that here: #4931
Closed it since it was only a draft, but can rebase it on top of this and see how it fares.
| // currentNumAnchorChans returns the current number of anchor channels the | ||
| // wallet should be ready to fee bump if needed. | ||
| // currentNumAnchorChans returns the current number of non-private anchor | ||
| // channels the wallet should be ready to fee bump if needed. |
There was a problem hiding this comment.
So this means from now on, since this will return 0 for private channels, we won't require private channels to reserver anything right?
| Anchors bool `long:"anchors" description:"enable support for anchor commitments"` | ||
| // NoAnchors should be set if we don't want to support opening or accepting | ||
| // channels having the anchor commitment type. | ||
| NoAnchors bool `long:"no-anchors" description:"disable support for anchor commitments"` |
There was a problem hiding this comment.
Just a thought, that it would be nice to not change this field but somehow leave it default to true.
There was a problem hiding this comment.
Unfortunately, IIRC, the flag parser we use cannot handle command line flags being false. So if we set this this true by default it would be no way of opting-out.
We cap the maximum value we'll reserve for anchor channel fee bumping at 10 times the per-channel amount such that nodes with a high number of channels don't have to keep around a very large amount for the unlikely scanario that they all close at the same time.
Since private channels (most likely) won't be used for routing other than as last hop, they bear a smaller risk that we must quickly force close them in order to resolve a HTLC up/downstream. In the case where it actually has to be force to resolve a payment (if it is the first/last hop on a routed payment), we can assume that the router will have UTXOs available from the reserved value from the incoming public channel.
This commit again enables anchor channels by default. During itests we still keep anchors opt-in, as many of the tests rely on the behaviour of non-anchor channels.
halseth
force-pushed
the
anchors-reserved-value-max
branch
from
0b2c7b1
to
820e77d
Compare
| // lnd, or to enable experimental protocol changes. | ||
| type ProtocolOptions struct { | ||
| // LegacyProtocol is a sub-config that houses all the legacy protocol | ||
| // options. These are mostly used for integration tests as most modern |
| type ProtocolOptions struct { | ||
| // LegacyProtocol is a sub-config that houses all the legacy protocol | ||
| // options. These are mostly used for integration tests as most modern | ||
| // nodes shuld always run with them on by default. |
There was a problem hiding this comment.
typo: shuld -> should
| for _, c := range chans { | ||
| cntChannel := func(c *channeldb.OpenChannel) { | ||
| // We skip private channels, as we assume they won't be used | ||
| // for routing. |
There was a problem hiding this comment.
I believe this assumption is incorrect because of non-strict forwarding. If one has a public channel and a private one with the same node, then it's possible to forward through both.
Disclaimer: I really don't understand anchor channels properly yet, so I may be wrong or this is non-issue.
We cap the maximum value we'll reserve for anchor channel fee bumping at
10 times the per-channel amount such that nodes with a high number of
channels don't have to keep around a very large amount for the unlikely
scanario that they all close at the same time.