Evidence-0003: Retention at Rank II

Report Date	2024/10/28
Submitted by	skunert

Member details

• Matrix username: @sebastian:parity.io
• Polkadot address: 1682A5hxfiS1Kn1jrUnMYv14T9EuEnsgnBbujGfYbeEbSK3w
• Current rank: 2
• Date of initial induction: 2024/01/08
• Date of last report: 2024/07/29
• Area(s) of Expertise/Interest: Cumulus Node, Substrate Node, Parachain Authoring, Weight

Reporting period

• Start date: 2024/07/30
• End date: 2024/10/30

Evidence

Proof weight hardening

During this reporting period I focused on the hardening of our proof weight system. Earlier this year we identified two issues with the way we handle proof size weight:

1. #5229: Since PoV-reclaim only monitors SignedExtensions post-dispatch hook, it was missing trie accesses that occured in pre-dispatch hooks. Together with unbenchmarked extrinsic base weights on parachains, this can lead to an underestimation of proof weight.
2. #6020: Both our benchmarking code and PoV-reclaim do currently not account for storage accesses that happen during the storage root calculation in finalize_block.

I worked on a quick fix to alleviate the impact of #5229 (#5281, #5257). However, proper hardening of the proof weight includes multiple steps, one of which is to integrate overhead benchmarking into the frame-omni-bencher in polkadot-sdk. Before, benchmarking block overhead and extrinsic base weight was not supported for parachains at all. I worked on identifying the requirements and came up with the design in #5303. Then I implemented the proposed changes, leading to #5891.

Slot-based collator

Another workstream has been the continuation of the slot-based collator improvements #6066. The MVP is in place and works for elastic scaling chains which want to produce blocks at fixed intervals, with one block per slot. However, the slot-based collator should aim to fully replace our existing collator implementations with the following improvements:

• Simplified modification of block production timings and conditions
• Introduction of a slot-based collator variant that is behaviour-compatible with the lookahead collator.
• Implementation of multi-block authoring support per parachain slot

Work on this has started but is not yet presentable.

Support

I invested time in supportive work:

• Sparred with Michal on the fork-aware transaction pool and spent a significant amount of time on its review #4639
• Reviewed fellowship RFC #103
• Reviewed runtime PRs (3) list
• Reviewed polkadot-sdk PRs (69) list

Voting record

Provide your voting record in relation to required thresholds for your rank.

Ranks	Activity thresholds	Agreement thresholds	Member's voting activities	Comments
III	70%	N/A
I	90%	N/A		There was nothing to vote on.
II	80%	N/A		There was nothing to vote on.
III	70%	N/A
IV	60%	N/A
V	50%	N/A
VI	40%	N/A

Misc

• Question(s):

• Concern(s):

• Comment(s):