I upgraded my node at coinos.io from bitcoin 25.0 to 28.0 the other day and got hit by a bunch of somewhat lengthy rescans that caused me about 30 minutes of downtime.
I have close to 200 watch-only descriptor wallets loaded at startup and each one of them had to go through about a 30-40s rescan as the node rebooted.
As this was occuring the node was unresponsive to any other RPC calls.
I attempted to abort some of the rescans with bitcoin-cli -rpcwallet= to no avail. I can’t recall if it gave me an error message or just did nothing.
I’m wondering if there is anything I can do to avoid this in the future. Were the rescans really necessary in the first place and is there any way to opt out of them, abort them, or improve the efficiency of them without missing incoming transactions or losing track of the balance of my wallets?
I anticipate adding significantly more watch-only wallets to the node in the future and was hoping to be able to manage thousands of them on a single machine. Is that unrealistic?
