1

After weeks of being in sync and working fine, my Geth just stopped. Upon restart, what started as a long hang on FS Sync Time (repeat output with different values, no further movement beyond allocation of cache and file handles), turned into a panic on latest launch:

INFO [03-17|12:07:14.213] Initialising Ethereum protocol           versions="[63 62]" network=1
WARN [03-17|12:07:15.325] Head block missing, resetting chain      hash=5414aa…aaf1f5
WARN [03-17|12:07:15.325] Rewinding blockchain                     target=0
panic: interface conversion: interface {} is nil, not *types.Block

goroutine 1 [running]:
github.com/ethereum/go-ethereum/core.(*BlockChain).CurrentBlock(...)
        /home/travis/gopath/src/github.com/ethereum/go-ethereum/core/blockchain.go:359
github.com/ethereum/go-ethereum/core.(*BlockChain).SetHead(0x400c4aa000, 0x0, 0x0, 0x0)
        /home/travis/gopath/src/github.com/ethereum/go-ethereum/core/blockchain.go:302 +0x8a0
github.com/ethereum/go-ethereum/core.(*BlockChain).ResetWithGenesisBlock(0x400c4aa000, 0x40002d4120, 0x0, 0x0)

I am now unable to progress past this point.

  • is your geth client up to date? – MSwezey Mar 17 at 19:18
  • Yes, 1.8.23 stable. – Swader Mar 17 at 19:25
0

Solution: remove ethereum files, and start geth again.

Ethereum Geth use "fast" option defaultly.

fast option defaulty have only 10,000 lastest blocks.

If the difference between the lastest block in public Ethereum and your lastest block on storage is over 10,000. "Head block missing" occurs.

Because fast option only needs to download recent 10,000 blocks and state.

If difference is big, users should download more than 10,000 blocks.

"Downloading more than 10,000 blocks" wastes user's resource.

  • I am running full nodes and have the whole 171GB of chaindata required. Starting from scratch is not an option because it would take another week to get in sync. – Swader Mar 17 at 18:15
  • what's your setup specs? I did a full, fast sync in 30 hours & 40 minutes – MSwezey Mar 17 at 19:19
  • NanoPC-T4. It's an experiment in nodes on microcomputers: reddit.com/r/ethereum/comments/alnmxn/… – Swader Mar 17 at 19:52
  • Yes, but there is no solution except for removing files. Archive node has the whole 2TB. – TLHBM Mar 18 at 4:43
  • Not sure what archive nodes have to do with anything, I'm running full nodes. Anyway, deleting everything is obviously not acceptable, not just in my case but in general. Not being able to continue from a non-zero block would make geth utterly broken software at this point in Ethereum's lifespan. I'm sure it's not that badly designed, otherwise no one would be running nodes. – Swader Mar 18 at 7:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.