# Classic Geth Nodes Syncing Slowly in ETC

Report by Yaz Khoury and Gregory Markou
## The Problem

### **Note: Everything in the ETC network is Fine!**
We have noticed that Classic Geth nodes are syncing slower than the other clients (Parity and Multi-Geth) in ETC.

We have isolated the problem to be the sweeper script running
for State Sweeping bloat for EIP-161 as found in this [repository](https://github.com/zmitton/statesweep)

We believe this script has slowed down Classic Geth syncing.
It could be an I/O amplification issue as pointed out by [Tomasz Zdybal](https://twitter.com/tzdybal).

Now that the State Sweep script has turned off, Classic Geth nodes
are syncing normally and catching up with the rest of the block head
on the other two clients. No network split
has occurred, just Classic-Geth nodes have
been slow to sync up.

This is most likely due to Classic Geth not being tested on state
sweep script while Parity and Multi-Geth have been with
Byzantium hard fork.

## State Sweep

State Sweep is for [EIP-161](https://github.com/ethereum/EIPs/blob/master/EIPS/eip-161.md), "State trie clearing (invariant-preserving alternative)" authored by Gavin Woods.

State trie clearing was to get rid of accounts with 0 balances
called empty accounts which had zero nonce and zero code which
were bloating up the network.

## Next Steps
We have turned off the state-sweep script for now which has resulted in Classic Geth nodes syncing normally again and catching up with the network.

If you're running Classic Geth nodes, it'll be syncing back normally again. We won't re-run the state sweep
script until we have had discussions with all the other clients about the next steps.