Lnd: LND fails after unlocking wallet (panic: runtime error: invalid memory address or nil pointer dereference)

[ManveerJarosz]

After I unlock my wallet, I get this error anytime I try to run a lncli command:

[lncli] rpc error: code = Unavailable desc = grpc: the connection is unavailable

This is what I'm seeing in my logs:

Mar 22 15:24:34 thundroid lnd[667]: 2018-03-22 15:24:34.423 [INF] CHDB: Checking for schema update: latest_version=0, db_version=0
Mar 22 15:24:34 thundroid lnd[667]: 2018-03-22 15:24:34.549 [INF] RPCS: password RPC server listening on 127.0.0.1:10009
Mar 22 15:24:34 thundroid lnd[667]: 2018-03-22 15:24:34.549 [INF] RPCS: password gRPC proxy started at 127.0.0.1:8080
Mar 22 15:24:34 thundroid lnd[667]: 2018-03-22 15:24:34.550 [INF] LTND: Waiting for wallet encryption password. Use lncli create to create wallet, or lncli unlock to unlock already created wallet.
Mar 22 15:32:59 thundroid lnd[667]: 2018-03-22 15:32:59.367 [INF] LNWL: Opened wallet
Mar 22 15:32:59 thundroid lnd[667]: 2018-03-22 15:32:59.536 [ERR] SRVR: invalid password
Mar 22 15:32:59 thundroid lnd[667]: panic: runtime error: invalid memory address or nil pointer dereference
Mar 22 15:32:59 thundroid systemd[1]: lnd.service: Main process exited, code=exited, status=2/INVALIDARGUMENT
Mar 22 15:32:59 thundroid systemd[1]: lnd.service: Unit entered failed state.
Mar 22 15:32:59 thundroid systemd[1]: lnd.service: Failed with result 'exit-code'.

I notice that the log says "invalid password", but in terminal it's showing:

Input wallet password:

lnd successfully unlocked!

Any idea what could have gone wrong?

I tried creating a new wallet password (but using the same seed that lnd gave me the first time), and this issue still persists.

[Roasbeef]

How much memory does what machine you're running on have? Does it have any swap?

[Roasbeef]

What's the full panic trace?

[ManveerJarosz]

Memory on my machine is 1994 MB, and only about 25% of it is being used.

I disabled Swap because the operating system is running from an SD card (the node is on an odroid).

I don't know what a full panic trace is, nor how I can retrieve it 😅

[Roasbeef]

Was this resolved?

[ManveerJarosz]

Yes :)

For anyone else who experiences this same issue: looks like this was caused by switching from testnet to mainnet. When I deleted macaroons.db, admin.macaroons, and readonly.macaroons, then no more errors after that.