Backup umbrel to NAS via autofs not working

Hello,
I got a problem with umbrel using autofs to backup the LND channel file to a NAS.
Autofs is also installed on a Pi3

PRETTY_NAME="Raspbian GNU/Linux 11 (bullseye)"
NAME="Raspbian GNU/Linux"
VERSION_ID="11"
VERSION="11 (bullseye)"
VERSION_CODENAME=bullseye
ID=raspbian
ID_LIKE=debian

and Umbrel + autofs is installed on a Pi4

PRETTY_NAME="Debian GNU/Linux 10 (buster)"
NAME="Debian GNU/Linux"
VERSION_ID="10"
VERSION="10 (buster)"
VERSION_CODENAME=buster
ID=debian

I copied the config files one by one from Pi3 to Pi4, but only Pi3 will work fine.

Umbrel debug log:

Starting automounter version 5.1.2, master map /etc/auto.master
using kernel protocol version 5.05
lookup(file): failed to read included master map auto.master
mounted indirect on /mnt/XXXX/ with timeout 60, freq 15 seconds
ghosting enabled
attempting to mount entry /mnt/XXXX/YYYY
mounted /mnt/XXXX/YYYY
re-reading map for /mnt/XXXX/
do_mount_indirect: indirect trigger not valid or already mounted /mnt/XXXX/YYYY
dev_ioctl_send_fail:499: AUTOFS_DEV_IOCTL_FAIL: error Invalid argument
do_mount_indirect: indirect trigger not valid or already mounted /mnt/XXXX/YYYY
dev_ioctl_send_fail:499: AUTOFS_DEV_IOCTL_FAIL: error Invalid argument
do_mount_indirect: indirect trigger not valid or already mounted /mnt/XXXX/YYYY
dev_ioctl_send_fail:499: AUTOFS_DEV_IOCTL_FAIL: error Invalid argument

while Pi3 does it right:

Starting automounter version 5.1.7, master map /etc/auto.master
using kernel protocol version 5.05
mounted indirect on /mnt/XXXX with timeout 60, freq 15 seconds
ghosting enabled
attempting to mount entry /mnt/XXXX/YYYY
mounted /mnt/XXXX/YYYY

The configs are:

root@umbrel:/# cat /etc/auto.master.d/syno.autofs
/mnt/XXXX/ /etc/auto.master.d/syno.map -t=60 --ghost

root@umbrel:/# cat /etc/auto.master.d/syno.map
YYYY -fstype=cifs,uid=1000,gid=1000,credentials=/etc/auto.master.d/synocred,iocharset=utf8,nosuid,nodev,vers=3.0 ://192.168.z.zzz/folder

Does somebody has any idea why Pi3 is running correctly while Umbrel is not?
Thanks :slight_smile:

The only file you can and you should make a rsync is channels.backup file located in
/home/umbrel/umbrel/lnd/data/chain/bitcoin/mainnet/channel.backup

That is the SCB backup that is updated every time you open or close a channel. And is all that matter + your node wallet seed.

Anything else is USELESS to do any rsync because is obsolete. It is changing ALL THE TIME.

If you really want a full backup in REAL TIME, use a RAID system, with 2 or 3 disks.
But with the channels.backup + seed is all you need to have your node recoverable. You can use optionally watchtowers also.

Don’t bother with that copy to NAS. Is useless.

How can I get this SCB if umbrel crashes suddenly?

And is this backup only updated by opened and closed channels?
I thought SCB gets an update every time LND funds will be transferred :hushed:
Because channel.backup is doing a force close on this channel, right? So it has to commit all transaction history, hasn’t it?

Depends what you mean by “crash suddenly”.
If is just a minor “not starting” doesn’t mean is crashed. You need to diagnose what is going one.
Jumping straight into “recovery” is not the first you should do. Is the last thing you should do.
That file can be accessed anytime, taking your drive and access it from another PC.

Yes.

Yes, standard procedure.
There are others ways, more advanced and require you to have knowledge about this stuff.
Read my guide and pay attention not just read the surface.
https://community.getumbrel.com/t/how-to-restore-your-node-wallet/

There is one thing left I don’t understand:
If I’m going to force close a channel, I have to commit an actual channel balance, haven’t I?
Or do I just request the peer node to close the channel by using channel.backup?
Because in first case I would have to save the channel.backup after every transaction to have an actual balance :thinking:

Thanks for helping :slightly_smiling_face:

No, SCB = static channel backup. That means is just a snapshot of the state of your channels.
Umbrel is doing that SCB EVERY time a channel is open or closed. There’s NO NEED to do it after each transaction.
All transactions and statuses are recorded in channels.db file (that is huge) and is written in real time.
Please start reading LND documentation how these things work.
https://wiki.ion.radar.tech/tutorials/troubleshooting/static-channel-backups

Also here a full list of documentation to be read:
https://community.getumbrel.com/t/important-links-to-learn-more-about-bitcoin-lightning-network/