Geth Fatal: Could not open database

[swlfigo]

Using Geth v1.3.1
Following Official Documentation:
https://docs.bnbchain.org/docs/validator/fullnode/

I download Geth fast node Snapshot from 48Club:
https://github.com/48Club/bsc-snapshots#geth-fast-node

First I create datadir using this command:
geth --datadir ./node --db.engine=pebble init ./genesis.json

then I replace donwloaded snapshot chaindata to replace folder in node's same place
./node/geth/chaindata

then I run Geth with this command:
geth --config ./config.toml --datadir ./node --cache 8000 --rpc.allow-unprotected-txs --txlookuplimit=0 --syncmode=full --tries-verify-mode=none --pruneancient=true --history.transactions=0 --db.engine=pebble

It shows error:

database=/node/geth/chaindata cache=3.12GiB 
databaseVersion: 8 (0x8)
headBlockHash: 0x753f3b10dc5400e0dd27e9470e5148ff557d3101a46a82724586838971fc1fcd
headFastBlockHash: x753f3b10dc5400e0dd27e9470e5148ff557d3101a46a82724586838971fc1fco
headHeaderHash: 0x753f3b10dc5400e0dd27e9470e5148ff557d3101a46a82724586838971fc1fcd
lastPivotNumber: <nil>
len(snapshotSyncStatus): o bytes
snapshotDisabled: false
snapshotJournal: 6355917 bytes
snapshotRecoveryNumber: <nil>
snapshotRoot: 0x660887bd6cddadb76041e13a5d21933a18c494104993ed2a1666de3cf63b2836
txIndexTail:  (x0)
fastTxLookupLimit: <nil>
Fatal: Could not open database: open /node/geth/chaindata/ancient/chain/bodies.cidx: no such file or directorv

I have tried backup files in ancient/chain after first initial with genesis.json, which it included database needed files, after move 48club snapshot chaindata to datadir, I replace original ancient/chain data in it. And Geth finally run and sync successfully

After successfully run geth , the file in node/geth/chaindata/ancient will lost and can't run next time

Is any wrong step I maked?

[RenRick]

”After successfully run geth , the file in node/geth/chaindata/ancient will lost and can't run next time“
Could you pls share the log? Or is it the same error with that you attached above? thanks

[joeylichang]

Step3：delete --pruneancient=true flag try again.
geth --config ./config.toml --datadir ./node --cache 8000 --rpc.allow-unprotected-txs --txlookuplimit=0 --syncmode=full --tries-verify-mode=none --history.transactions=0 --db.engine=pebble

note: don't replace snapshot's ancient/chain data.

[swlfigo]

Step3：delete --pruneancient=true flag try again. geth --config ./config.toml --datadir ./node --cache 8000 --rpc.allow-unprotected-txs --txlookuplimit=0 --syncmode=full --tries-verify-mode=none --history.transactions=0 --db.engine=pebble

note: don't replace snapshot's ancient/chain data.

@joeylichang The snapshot downloaded from 48club, there is no xxxxx.cidx files in snapshot/geth/chaindata/ancient/ folder, if I don't replace folder created with genesis.json , Geth will show can't open DB which can't find these files.

After Delete --pruneancient=true flag , same error . the main reason is missing xxxx files which open db needed

[swlfigo]

”After successfully run geth , the file in node/geth/chaindata/ancient will lost and can't run next time“ Could you pls share the log? Or is it the same error with that you attached above? thanks

@RenRick same error, After First Success run, all file (eg. xxxx.cidx ) all missed in ancient folder.

[weiihann]

This should be due to the pruneancient bug, which has been fixed in this PR #1986.

[flywukong]

Compile geth with v1.3.1 + this PR #1986, and try run with the same snapshot and the commands as the issue descriped, it sync successfully .

t=2023-11-22T06:13:17+0000 lvl=info msg="Block synchronisation started"
t=2023-11-22T06:13:23+0000 lvl=info msg="Importing sidechain segment"           start=33,660,920 end=33,661,091
t=2023-11-22T06:13:24+0000 lvl=warn msg="Inserted known block"                  number=33,660,923 hash=0xec091fe0ebdfa5d2551bf5ad34a30be0e7e75ce48f7eb34beedd7b02b4e116a5 uncles=0 txs=0 gas=0 root=0x18daf402293b4b25a0f1f7f167c49e891e5f42479607d54a0989085528f560e9
t=2023-11-22T06:091fe0ebdfa5d2551bf5ad34a30be0e7e75ce48f7eb34beedd7b02b4e116a5 oldblocks=166 newnum=33,660,923 newhash=0xec091fe0ebdfa5d2551bf5ad34a30be0e7e75ce48f7eb34beedd7b02b4e116a5 newblocks=0
t=2023-11-22T06:13:31+0000 lvl=info msg="Imported new chain segment"             number=33,660,934 hash=0x62a0dd69a3943b60727605c76fbf1d5cea8e3ce2c0a57821f087cfbfda237ff1 miner=0xCc8E6d00C17eB431350C6c50d8b8F05176b90b11 blocks=15 txs=2313 mgas=286.555 elapsed=8.412s mgasps=34.065 age=1d15h6m  triedirty="0.00 B"13:25+0000 lvl=eror msg="Impossible reorg, please file an issue" oldnum=33,660,923 oldhash=0xec091fe0ebdfa5d2551bf5ad34a30be0e7e75ce48f7eb34beedd7b02b4e116a5 oldblocks=166 newnum=33,660,923 newhash=0xec091fe0ebdfa5d2551bf5ad34a30be0e7e75ce48f7eb34beedd7b02b4e116a5 newblocks=0
t=2023-11-22T06:13:31+0000 lvl=info msg="Imported new chain segment"             number=33,660,934 hash=0x62a0dd69a3943b60727605c76fbf1d5cea8e3ce2c0a57821f087cfbfda237ff1 miner=0xCc8E6d00C17eB431350C6c50d8b8F05176b90b11 blocks=15 txs=2313 mgas=286.555 elapsed=8.412s mgasps=34.065 age=1d15h6m  triedirty="0.00 B"

Pls notice it is not needed to replace original ancient/chain data after fixed now.

[swlfigo]

This should be due to the pruneancient bug, which has been fixed in this PR #1986.

problem is fixed with this PR, Thx!

[swlfigo]

problem is fixed with this PR, Thx!

problem is fixed with this PR, Thx!