Error starting macaroon service

[andrerfneves]

Hey there,

Weird start up issue, my macaroon path and TLS cert path are correct. I also think 127.0.0.1 is set correctly as we're testing on local LND. What is the reason for being unable to start the macaroon service? Running v0.12.1 with correct tags.

2021-03-06 22:32:55.295 [INF] LNDC: Creating lnd connection to 127.0.0.1:10009
2021-03-06 22:32:55.295 [INF] LNDC: Connected to lnd
2021-03-06 22:32:55.295 [INF] LNDC: Waiting for lnd to unlock
2021-03-06 22:32:55.356 [INF] LNDC: lnd version: v0.12.1-beta, build tags 'autopilotrpc,chainrpc,experimental,invoicesrpc,routerrpc,signrpc,walletrpc,watchtowerrpc,wtclientrpc'
2021-03-06 22:32:55.356 [INF] LNDC: Using network mainnet
2021-03-06 22:32:55.356 [INF] LNDC: Waiting for lnd to be fully synced to its chain backend, this might take a while
2021-03-06 22:32:55.414 [INF] LNDC: lnd is now fully synced to its chain backend
2021-03-06 22:33:00.379 [ERR] LITD: Could not start subservers: error starting macaroon service: unable to set up macaroon authentication: timeout
2021-03-06 22:33:00.379 [INF] SGNL: Received shutdown request.
error starting macaroon service: unable to set up macaroon authentication: timeout
2021-03-06 22:33:00.379 [INF] SGNL: Shutting down...

Thanks!

[jamaljsr]

Hey @andrerfneves which version of LiT are you running and what flags are you using in the lit.conf or command line?

[guggero]

I think it's trying to open the macaroon database but runs into a timeout because it's already open. Are you sure there's no instance of Faraday, Loop or Pool running at the same time as you run LiT? LiT by default starts all three mentioned daemons and they all require exclusive lock on their respective databases. If you want to continue using the daemons you already run, use --<daemon>-mode=remote that was added in the latest version.

[andrerfneves]

Thanks @jamaljsr and @guggero. The issue was I had another Faraday service auto-restarting after I killed it to run LiT's builds. Thanks for the help. I should probably open another ticket, but any Loop In or Out is being returned with this alert:

Don't think I'm doing anything fancy, just followed the walkthrough steps.

[guggero]

Something seems to time out now. Are you running on Tor only? The added latency of Tor might be too much for Loop.

[andrerfneves]

No tor used/installed, clearnet only

[guggero]

Weird. Do you see this for every attempt or just sometimes?

[andrerfneves]

I've just spent some time trying to loop in or out with selected channels, or just automatically letting LiT select the channels to loop. All have so far lead to this deadline exceeded error.

[andrerfneves]

Is there else anything I can try to test out the part that seems to be failing? Or are we unclear right now as to what is timing out? I can dig deeper just need some leads

[jamaljsr]

Hey @andrerfneves are there any clues in the logs? Would you mind sharing a snippet from when the loop is initiated?

[andrerfneves]

@jamaljsr this is all there is. Maybe there's a loglevel?

2021-03-10 01:49:13.341 [INF] LOOP: Loop in 0.XXXX BTC (last hop: 03864ef025fde8fb587d989186ce6a4a186895ee44a926bfc370e2c366597a3f8f)
2021-03-10 01:49:13.701 [INF] LOOP: Creating probe invoice aa627569bbc304d9c7fb30fd7a4ec62e69c4cdf6bdaeb0d531502ce3b978e3c0
2021-03-10 01:49:14.010 [INF] LOOP: Initiating swap request at height 673930
2021-03-10 01:49:44.029 [ERR] LOOPD: Loop in: cannot initiate swap: rpc error: code = DeadlineExceeded desc = context deadline exceeded

[jamaljsr]

Thanks for sharing. The only thing that stands out to me is that loopd is not getting a response from the swap server in 30 seconds. I don't believe changing the loglevel would add any additional information to the logs.

Have you tried using the CLI? I don't expect the results to be different, but just to ensure that the UI isn't sending the wrong data, it ay be worth a try. Since you're running with loop integrated, you'll need to provide these flags to the loop CLI

loop --rpcserver=localhost:8443 --tlscertpath=~/.lit/tls.cert --macaroonpath=~/.loop/mainnet/loop.macaroon loop in ...

[guggero]

I looked at the server log for this swap and found the following:

2021-03-10 01:49:44.047 [ERR] NAUT: Swap cbdxxxxxxx initiation error: FailedPrecondition (target unreachable)

This means the payment probe we do for Loop In failed (not being able to find a path to your node). But the error you get is different so we probably swallow the correct error message somewhere.

I'm going to open an issue in the Loop repo instead and closing this one. Thanks for your input @andrerfneves!