-
Notifications
You must be signed in to change notification settings - Fork 1.6k
Req/res optimization for statement distribution #2803
Conversation
…des. (#2778) * Better timeout values. * Fix typo. * Fix validator bandwidth. * Fix compilation.
Most importantly code size is now 5 Meg, which is the limit we currently want to support in statement distribution.
@eskimor triage? |
Now that max code size is compressed size.
I am afraid I hurried that part a bit too much - likely not. Will check & fix. |
instead of relying on signatures alone.
And add test for this.
/// nodes within two hops, we will take about 2 seconds for transferring statements (data transfer | ||
/// only). If necessary, we could be able to reduce this to 3 seconds. To consider: The lower the | ||
/// riskier that we will not be able to include a candidate. | ||
const PROPOSE_TIMEOUT: core::time::Duration = core::time::Duration::from_millis(4000); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
What needs to happen after that 4 seconds and can it fit within the remaining two seconds?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Invoking the on_initialize
, adding transactions, and then on_finalize
of the block. 2 seconds is a bit of a squeeze, honestly.
When this timeout is hit, it means that no parachain stuff will be included which probably makes the 2 seconds more acceptable.
/// Maximum compressed code size we support right now. | ||
/// At the moment we have runtime upgrade on chain, which restricts scalability severly. If we want | ||
/// to have bigger values, we should fix that first. | ||
pub const MAX_CODE_SIZE: u32 = 3 * 1024 * 1024; | ||
|
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@rphmeier what do you think about this? I guess it'd have to be the upper limit for every chain that will have parachains?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'd suggest making it equal to whatever we have in #2854
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
polkadot/node/service/src/chain_spec.rs
Line 898 in 0f3a21c
max_code_size: MAX_CODE_SIZE, |
bot merge |
Waiting for commit status. |
Merge aborted: Checks failed for e506d64 |
bot merge |
Waiting for commit status. |
Status: Done. I am still working on some improvements, but those should not be blocking and can go in as a separate PR.