From 3de79a43b1666e4a6bdff19350013854ef94da0d Mon Sep 17 00:00:00 2001 From: alexander-sei Date: Fri, 20 Nov 2026 18:08:49 +0100 Subject: [PATCH] docs(node): warn that snapshots require trusting the provider Restoring a provider snapshot trusts the provider: a malicious or compromised provider can corrupt it in ways that are not immediately detectable, and the node cannot verify the historical state in it. State sync is the only officially supported bootstrap method for Sei Mainnet. Add that warning to the top of the snapshot sync guide, and stop the AppHash recovery steps in troubleshooting from calling snapshot providers trusted. Co-authored-by: Cursor --- node/snapshot.mdx | 2 ++ node/troubleshooting.mdx | 4 ++-- 2 files changed, 4 insertions(+), 2 deletions(-) diff --git a/node/snapshot.mdx b/node/snapshot.mdx index 129722a1..2f9b513f 100644 --- a/node/snapshot.mdx +++ b/node/snapshot.mdx @@ -4,6 +4,8 @@ sidebarTitle: 'Snapshot Sync' description: 'Detailed guide for using snapshots to join the network' keywords: ['sei node', 'blockchain node', 'node configuration', 'seidb', 'node maintenance', 'system operations'] --- +**Restoring a snapshot requires you to trust the snapshot provider.** A malicious or compromised provider could corrupt a snapshot in ways that are not immediately detectable. Your node cannot verify the historical state in a snapshot at all. For safe bootstrapping, use [state sync](/node/statesync). It is the only officially supported bootstrap method for Sei Mainnet. + Before you resync a node from a snapshot, make sure that your validator cannot double sign. After a successful resync, your validator must _under no circumstance_ double sign blocks at previous heights. If it does, the validator is tombstoned. Use snapshot sync to start a new full node and join an existing network quickly. Snapshot sync restores the node from a snapshot instead of replaying all historical blocks. diff --git a/node/troubleshooting.mdx b/node/troubleshooting.mdx index 69ef9190..b3d0b50c 100644 --- a/node/troubleshooting.mdx +++ b/node/troubleshooting.mdx @@ -204,9 +204,9 @@ block_id={"hash":"...","parts":{"hash":"...","total":1}} height=[HEIGHT] 1. **Stop the node immediately.** 2. **Try a node rollback first.** See [Node rollback](/node/troubleshooting#node-rollback). -3. **If the rollback fails, restore from a fresh snapshot:** +3. **If the rollback fails, resync the node:** - - Download a recent snapshot from trusted providers (Polkachu, PublicNode) + - Use [state sync](/node/statesync), or a recent [snapshot](/node/snapshot) from a provider that you trust - Make sure that you use the correct node version - Verify that peer configurations are up to date