Don't block on channel arbitator startup#9324
Conversation
|
Important Review skippedAuto reviews are limited to specific labels. 🏷️ Labels to auto review (1)
Please check the settings in the CodeRabbit UI or the You can disable this status message by setting the Thank you for using CodeRabbit. We offer it for free to the OSS community and would appreciate your support in helping us grow. If you find it useful, would you consider giving us a shout-out on your favorite social media? 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments. CodeRabbit Commands (Invoked using PR comments)
Other keywords and placeholders
CodeRabbit Configuration File (
|
2f28012 to
c1dc62b
Compare
guggero
left a comment
There was a problem hiding this comment.
Thanks a lot for looking into this. I think we can simplify the code somewhat and also reduce the number of goroutines a bit, see inline comment.
yyforyongyu
left a comment
There was a problem hiding this comment.
Left some comments here, I don't think the fix should happen in lnd. From the logs, what happened was the chan newInputs was blocked since we sent two inputs during the startup, and the first input was blocking on aux.ExtraBudgetForInputs,
2024-11-29 10:15:07.185 [INF] SWPR: Sweep request received: out_point=e4a01568a0cd879df94cd6c9ae8143e1304bf1731ddee7003c6f12df4f7baf13:0, witness_type=TaprootAnchorSweepSpend, relative_time_lock=0, absolute_time_lock=0, amount=0.00000330 BTC, parent=(<nil>),
...
2024-11-29 10:15:07.192 [INF] CNCT: ChannelArbitrator(ce3ab677a419e40558d095faf7165ba143467c1dc5691632ad511224473eabc9:0): offering anchor from local commitment 7a4c66554f168a903202ed29e2a13e1061b8c0f484a4be7a2ef8d69ceac382c5:1 to sweeper with deadline=None, budget=0.00000330 BTC
2024-11-29 10:15:07.192 [INF] SWPR: Sweep request received: out_point=7a4c66554f168a903202ed29e2a13e1061b8c0f484a4be7a2ef8d69ceac382c5:1, witness_type=TaprootAnchorSweepSpend, relative_time_lock=0, absolute_time_lock=0, amount=0.00000330 BTC, parent=(fee=0.00001366 BTC, weight=958 wu), params=(startingFeeRate={false 0}, immediate=false, exclusive_group=948947804208955392, budget=0.00000330 BTC, deadline=none)
There are multiple ways we can fix this,
- make
newInputsa buffered chan, which will defeatblockbeatas we no longer guarantee the orderchannel arbitrator offers -> sweeper sweeps. - we could instead remove the
advanceStatein theChannelArbitrator.Startsince the channels will be handled in the main goroutinechannelAttendantanyway. - fix it in the
tapdside, which feels more appropriate as if the calls to methods of theAuxSweeperare blocking, we may have had some incorrect assumptions when using them in thesweeper.
yyforyongyu
left a comment
There was a problem hiding this comment.
I tested out the second idea here, turns out it's a small change, and all the itests passed!
c1dc62b to
7efaffe
Compare
ziggie1984
left a comment
There was a problem hiding this comment.
Have some questions in terms of this new introduced behaviour, we are altering the lnd daemon logic slightly, which I think is ok and actually a better way to go instead of shutting down LND in case the startup of one arbitrator fails.
| if err != nil { | ||
| log.Errorf("Unable to progress state machine after "+ | ||
| "restart: %v", err) | ||
| return | ||
| } |
There was a problem hiding this comment.
This is a code change behaviour:
Previously we would not startup if the broadcast of the channel commitment failed, now we are just logging an error but we still continue running the daemon.
There was a problem hiding this comment.
If we want to restore the behavior, we can use Criticalf here.
There was a problem hiding this comment.
I kinda like this new approach, so if everybody is ok with it, I would favour not shutting down if one arbitrator fails
| if err != nil { | ||
| log.Errorf("Unable to progress state machine after "+ | ||
| "restart: %v", err) | ||
| return | ||
| } |
There was a problem hiding this comment.
If we want to restore the behavior, we can use Criticalf here.
514f29e to
2905de9
Compare
|
CI passed in the previous run, so I think this is gtg as soon as we have consensus whether we shutdown LND if a channel arbitrator fails to start or we just log an error and continue the process nonetheless. |
yyforyongyu
left a comment
There was a problem hiding this comment.
LGTM🦾 Thanks for the quick fix!
2905de9 to
8235f83
Compare
guggero
left a comment
There was a problem hiding this comment.
Nice, very clean solution, LGTM 🎉
We decouple the state machine of the channel arbitrator from the start-up process so that we can startup the whole daemon reliably.
8235f83 to
0004e31
Compare
Replaces #9262
Fixes: #9323