ERROR next_uncached_key_range went backwards

[MrFrogoz]

Describe the bug

Hi

Updating the node to the latest version I get the following error, the node automatically reboots and crashes with the same error

Steps to reproduce

Temp FIX
Starting the nodes with "--engine.legacy-state-root" ( the legacy engine ), the nodes started successfully

Node logs

reth[2129405]: 2026-03-12T00:29:21.343568Z INFO Stage finished pipeline_stages=13/13 stage=Finish checkpoint=24637710 target=24637710
reth[2129405]: 2026-03-12T00:29:21.872625Z ERROR next_uncached_key_range went backwards: calc_lower=Nibbles(0x0c2) < prev_lower=Nibbles(0x0c3), calc_upper=Some(Nibbles(0x0c3)), prefix=Nibbles(0x)
reth[2129405]: Thread 'proof-acct-08' (2129580) panicked at /home/runner/.rustup/toolchains/stable-x86_64-unknown-linux-gnu/lib/rustlib/src/rust/library/alloc/src/vec/mod.rs:2873:36:
reth[2129405]: Range start index 18446744073709551615 out of range for a slice of length 15
reth[2129405]: Stack backtrace:
reth[2129405]: Note: Some details were omitted, run RUST_BACKTRACE=full for a detailed backtrace.
reth[2129405]: Rayon: Unexpected panic error detected; aborting

Platform(s)

Linux (x86)

Container Type

Not running in a container

What version/commit are you on?

Reth Version: 1.11.2
Commit SHA: 793a3d5
Build Timestamp: 2026-03-10T11:50:47.051354330Z
Build Features: asm_keccak,jemalloc,keccak_cache_global,min_debug_logs,otlp,otlp_logs,rocksdb
Build Profile: maxperf

What database version are you on?

.

Which chain / network are you on?

ethereum mainnet

What type of node are you running?

Archive (default)

What prune config do you use, if any?

No response

If you've built Reth from source, provide the full command you used

No response

Code of Conduct

• I agree to follow the Code of Conduct

[b0a7]

I'm also seeing this

[MrFrogoz]

@b0a7 add this "--engine.legacy-state-root" until fix

[trial123Zel]

I received the same Nibble error while on v1.11.1 so I updated to v1.11.2 after seeing the changelog but I also received the same issue. I can confirm that the temporary fix of adding --engine.legacy-state-root solves the problem for now. Adding flag --engine.disable-parallel-sparse-trie did not solve the issue.