[Historical testnet] Fresh v0.1.0 nodes cannot sync: HistoryStep matrix whitelist slot 0 drift #5
Labels
No labels
bug
documentation
duplicate
enhancement
good first issue
help wanted
invalid
pool
post-quantum
question
third-party
wontfix
No milestone
No project
No assignees
1 participant
Notifications
Due date
No due date set.
Dependencies
No dependencies set.
Reference
ignotusnemo/parano1d#5
Loading…
Add table
Add a link
Reference in a new issue
No description provided.
Delete branch "%!s()"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
Fresh v0.1.0 mainnet nodes are unable to sync from height 0.
The issue reproduces on multiple independent x86_64 machines.
P2P connectivity works correctly. The nodes connect to multiple peers and receive state manifests from peers with current chain tips, but the manifests are rejected during HistoryStep verification.
Example error:
BadHistoryStepTerminal(
"HistoryStep terminal rejected: HistoryStep bank:
HistoryStep matrix whitelist slot 0 drift"
)
The node then reports:
state manifest round produced no usable candidate — re-requesting
I have also observed:
HistoryStep proof: Zerocheck(SumcheckFinalFailed)
One affected machine:
The same sync failure also reproduces on another x86_64 machine with 32 threads.
On the EPYC node, peers are discovered successfully and manifests are received, for example:
received state manifest ... tip=133 segments=1
but then:
ignoring manifest from a peer that supplied an invalid recursive terminal
The node remains at:
Height 0
while connected peers advertise chain heights well above 100.
Steps to reproduce:
./parano1d --check-hardware
Result:
BACKEND avx512bw+vpclmul
NODE READY
./parano1d
Peers connect successfully, but the node remains at height 0 and cannot accept a state snapshot.
Running with:
./parano1d --log debug
shows the HistoryStep terminal verification failure described above.
This appears reproducible across different hardware, so it does not seem specific to the EPYC system.
Confirmed. Thank you for the detailed report and logs.
This is not hardware-specific and is not caused by the AVX-512 backend. The public v0.1.0 binaries were built by CI with the pre-mainnet test HistoryStep pack, while the final mainnet pack was used in the local build tested before release. That mismatch causes both errors reported here:
HistoryStep matrix whitelist slot 0 driftHistoryStep proof: Zerocheck(SumcheckFinalFailed)The source code, genesis and mainnet parameters are correct. The published v0.1.0 binaries are not and must not be used.
I have hidden v0.1.0 and stopped all seed nodes. v0.1.1 is currently being built for all supported platforms with the final mainnet HistoryStep pack.
Once v0.1.1 is published, affected users must close v0.1.0, delete the complete Parano1d data directory, install v0.1.1 and start again from a clean directory:
~/.parano1d/data/Users/YourName/.parano1d/data%USERPROFILE%\.parano1d\dataI will keep this issue open until v0.1.1 is published, all seed nodes are restarted, and clean public synchronization is verified.
Historical testnet issue.
This report concerns the retired pre-mainnet network and its early v0.1.0/v2.x binaries. It does not apply to the current Parano1d mainnet, which launched from a new genesis on August 21, 2026.
The original failure was caused by incompatible HistoryStep packs in early CI and local builds. Current mainnet binaries use the canonical authenticated pack and bind its identity before P2P synchronization, preventing incompatible packs from joining the same network.
The networking redesign is documented in PR #6.
Current mainnet release: https://github.com/ignotusnemo/parano1d/releases/latest