I was running v4.1.1 (upgraded days ago), previously running v4.1.0 with no problems.
Reindexing now to recover & preserved debug.log just in case.
EDIT: Probably a red herring, IP addr (external) changed after the node was started but before it finished syncing.
EDIT: It just died during the reindex (at block 230035), here’s the error :-
zcashd: main.cpp:2597: DisconnectResult DisconnectBlock(const CBlock &, CValidationState &, const CBlockIndex *, CCoinsViewCache &, const CChainParams &, const bool): Assertion `pindex->nCachedBranchId’ failed.
Aborted
EDIT: Building a new node from scratch & will import the old wallet.dat (its gonna take a while)
EDIT: 12 hrs later & its at 18%, really should upgrade that machine…
EDIT: Syncing during a total solar eclipse, this is gonna be a very special node…19%…
EDIT: 365GB total RX & 48Mb/s RX (according to iftop), at block 212653 (19%) after 20 hrs
That’s a LOT of data, very different to the last time I synced a node from scratch!!!
Attaching a screenshot showing the node console, iftop, debug.log & top :-
EDIT:
I’m calling it ‘officially pooched’ & aborting the sync.
The node had been running for 24 hours, it synced to block 217784 & used 440GB download - something is broken.
EDIT:
Just for the sake of completeness I restarted the node, it resumed syncing at a MUCH faster rate.
I’ll leave it running, but for now I would say ‘initial sync is pooched’, ‘resumed/restarted sync works’.
EDIT: Confirmed. Its been 2hrs since restarting the node & its synced to block 298300, much faster.
EDIT: Checked it again this morning & it’d bogged down, syncing slowly & had reached block 373610. Restarted the node & it was much faster.
Addtional: Noticed warnings in debug.log ‘overwinter version too high’, which is new to me.
EDIT: Restarted with ‘-connect=node.IP.address.here’ so it could sync from another local node, much MUCH faster & will let it sync to completion.
EDIT: I’m seeing BAN messages in debug.log ‘Misbehaving: BAN THRESHOLD EXCEEDED’, the node its syncing from is running v4.1.1. I’m using the ‘-connect’ option so the node just reconnects automatically & continues, but the ban shouldn’t be happening.
Here’s what the misbehaving is about from debug.log :-
Dec 15 11:31:29.197 ERROR main: CScriptCheck(): d493018c4c53af2825d54560108d5e5523d7ebb6a89098f2af1691ef8de1ef0a:0 VerifySignature failed: Script evaluated without error but finished with a false/empty top stack element
Dec 15 11:31:29.197 ERROR main: AcceptToMemoryPool: ConnectInputs failed d493018c4c53af2825d54560108d5e5523d7ebb6a89098f2af1691ef8de1ef0a
Dec 15 11:31:29.197 INFO main: Misbehaving: 192.168.0.71 (0 → 100) BAN THRESHOLD EXCEEDED
EDIT: Looks to me like the syncing node is seeing forwarded Canopy TXNs as invalid as it hasn’t yet reached the activation block (1046400), which is causing it to flag the sending node as ‘naughty’ and after 100 txns received it issues a ban.
SUMMARY: The node is back, synced, old wallet installed and rescanned.
The CopyPreviousWitnesses problem was probably caused by ‘bad things’ happening to this machine, its been killed several times, power outages, all sorts. I can easily imagine some disc corruption caused by that.
The Initial Sync problem, is a real problem - one I think we’ve seen before ? @LarryRuane