MemTableSize with Pebble

[quickchase]

System information

Geth version:

Version: 1.11.2-stable
Git Commit: 73b01f40ceb6bcb6f9f44c2a3d6f963b40452b47
Git Commit Date: 20230222
Architecture: amd64
Go Version: go1.20.1

CL client & version:

Lighthouse v3.5.0-0fb58a6
BLS library: blst-modern
SHA256 hardware acceleration: true
Allocator: jemalloc
Specs: mainnet (true), minimal (false), gnosis (true)

OS & Version:

Ubuntu 20.04.5 LTS (GNU/Linux 5.15.0-1029-oracle x86_64)

I'm trying to understand the following error:

Started Geth.
INFO [02-28|02:38:20.737] Starting pprof server                    addr=http://127.0.0.1:6061/debug/pprof
INFO [02-28|02:38:20.743] Starting Geth on Ethereum mainnet...
INFO [02-28|02:38:20.744] Maximum peer count                       ETH=50 LES=0 total=50
INFO [02-28|02:38:20.745] Smartcard socket not found, disabling    err="stat /run/pcscd/pcscd.comm: no such file or directory"
INFO [02-28|02:38:20.745] Using pebble as db engine
WARN [02-28|02:38:20.747] Sanitizing cache to Go's GC limits       provided=64408 updated=42938
INFO [02-28|02:38:20.749] Set global gas cap                       cap=600,000,000
INFO [02-28|02:38:20.761] Allocated trie memory caches             clean=6.29GiB dirty=10.48GiB
INFO [02-28|02:38:20.761] Using pebble as the backing database
INFO [02-28|02:38:20.761] Allocated cache and file handles         database=/home/eth/geth/geth/chaindata cache=20.97GiB handles=524,288
Fatal: Failed to register the Ethereum service: MemTableSize (4.0 G) must be < 4.0 G

The only reference to MemTableSize I could find was #24615, which talks about 64-bit vs 32-bit etc. but the system/VM here is a 64-bit VM

I don't know where this setting comes from, how to set it, or what to do about it.

I have a different Oracle VM with the same settings, and it worked fine. The only real difference I saw between them was a kernel version:

5.13.0-1018-oracle

Any guidance on where to look to solve this issue would be fantastic...

[quickchase]

Ah ha! It has to do with the --cache setting! Because the other VM has less RAM the cache was configured to a lower value, but this system had more RAM, thus the cache setting was set higher...causing this error.

The simple solution is to use a more sane cache value in my config 😆

[quickchase]

Oh I suppose that would fix it as well 🙂