Does db snapshots work for cumulus collators?

[samelamin]

Looking at the code it seems like snapshots cannot be supported because we are writing the snapshot file to the chains dir but for collators, we also have the relay-data directory. A snapshot for a parachain needs both directories

https://github.com/paritytech/zombienet/blob/HEAD/javascript/packages/orchestrator/src/providers/native/nativeClient.ts#L283-L284

https://github1s.com/paritytech/zombienet/blob/HEAD/javascript/packages/orchestrator/src/providers/podman/podmanClient.ts#L383-L384

[Relaychain] Bootnode with peer id `12D3KooWQCkBm1BYtkHpocxCwMgR8yjitEeHGx8spzcDLGt2gkBm` is on a different chain (our genesis: 0x2d46…5d89 theirs: 0x2540…b2b2)  

[pepoviola]

Hi @samelamin, thanks for reporting this issue and provide feedback. I think the simplest solution is to match the node role to extract the data in the correct directory.
Thanks!

[samelamin]

No worries @pepoviola I have sent a pr so feel free to review it when you get a chance.

Wouldnt it be simpler to extract both at the root? because otherwise you need to match the node role as well as prepare the snapshot differently for both relay and parachain.

While if we extract at the root, everything remains consistent?

[pepoviola]

Hi @samelamin, thanks for the pr. I would review and check it asap.
Thanks!!

cc: @michalkucharczyk if we change this we will need to update the CI test for substrate.

[pepoviola]

Hi @samelamin, closing here since the pr is already merged and included in the latest version.
Thanks!