our team has tested our own Relay Chain(forked from polkadot-release-1.10) with async backing feature based on the wiki. Our Relay Chain has two operating parachains. However, it seems like Relay Chain's pvc size of parachain data is growing so fast and found out it was due to relay_storage_proof, which is relay_dispatch_queue_remaining_capacity(roughly 1.6mb). We don't know where to fix this problem. Thank you
1 Answer
Correct me if I'm wrong.
I think it might be caused by this:
/// We allow for 2 seconds of compute with a 6 second average block.
pub const MAXIMUM_BLOCK_WEIGHT: Weight = Weight::from_parts(
WEIGHT_REF_TIME_PER_SECOND.saturating_mul(2),
cumulus_primitives_core::relay_chain::MAX_POV_SIZE as u64,
);
For our parachain, this increases our block space fourfold.
-
Thank you for your reply. I think it is not related to size of parachain's blockspace. The problem is size of the relay chain state proof size while constructing parachain inherent data. We still haven't found out..– RafCommented Jul 17 at 5:08
polkadot-sdk. Please include as much information as possible. Do you mean that callingrelay_dispatch_queue_remaining_capacityis increasing the proof size that much?state_version=0in Runtime which should be1. By the way, what is thisstate_versionused for?nodeorleafanymore. Thus, it also explains why it solves your issue here :) When settingstate_version = 1this new format is used. Good that you solved it! The production networks Polkadot/Kusama using this also now :)