-
Notifications
You must be signed in to change notification settings - Fork 699
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Add simple collator election mechanism #1340
Conversation
Signed-off-by: georgepisaltu <george.pisaltu@parity.io>
Signed-off-by: georgepisaltu <george.pisaltu@parity.io>
Signed-off-by: georgepisaltu <george.pisaltu@parity.io>
Can I get a brief summary of why |
} | ||
|
||
/// The caller `origin` replaces a candidate `target` in the collator | ||
/// candidate list by reserving `deposit`. The amount `deposit` |
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.
line width is not according to our documentation guideline, please use exactly the correct line-width, not less.
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.
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.
Fixed.
Long discussion here, but in short the benchmarks suggest In the previous PR linked above I had POCs with |
idx += 1; | ||
while idx < candidate_count && candidates[idx].deposit < new_deposit { | ||
candidates.as_mut().swap(idx - 1, idx); | ||
idx += 1; | ||
} | ||
} else { | ||
while idx > 0 && candidates[idx].deposit >= new_deposit { | ||
candidates.as_mut().swap(idx - 1, idx); | ||
idx -= 1; |
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 think this is safe as long as candidates
is less than u32::MAX
, but not sure if it's still FRAME best practice to use something like saturating_accrue
. @gpestana ?
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.
The length of candidates
can be at most u32::MAX
, since the BoundedVec
is capped by MaxCandidates
, which is Get<u32>
. However, idx
must be less than u32::MAX
. When creating idx
through candidates.iter().position()
, if a match is found, it will be at most u32::MAX - 1
even if it's the last element of a u32::MAX
-sized list. As things stand, no operation above should overflow or underflow.
idx += 1; | ||
while idx < list.len() && list[idx].deposit < deposit { | ||
list.as_mut().swap(idx - 1, idx); | ||
idx += 1; |
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 think you should use checked_add
here and return an Overflow
error to prevent any state mutations.
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.
The same assumptions mentioned previously about the list being capped by MaxCandidates: Get<u32>
and the idx
of elements still hold here. Would it be ok to leave as is if there were UTs targeting specifically these potential overflows?
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 think so but prefer someone else on FRAME to comment.
Signed-off-by: georgepisaltu <george.pisaltu@parity.io>
Signed-off-by: georgepisaltu <george.pisaltu@parity.io>
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.
Few comments but overall good. Still prefer safe math in the runtime even if it's paranoid.
Signed-off-by: georgepisaltu <george.pisaltu@parity.io>
Signed-off-by: georgepisaltu <george.pisaltu@parity.io>
Signed-off-by: georgepisaltu <george.pisaltu@parity.io>
* ignore errors when dumping logs and container is missing * fixed typo * print correct payload length * increase conversion rate a bit when estimating fee (to avoid message rejects when rate update tx is active) * fmt
* ignore errors when dumping logs and container is missing * fixed typo * print correct payload length * increase conversion rate a bit when estimating fee (to avoid message rejects when rate update tx is active) * fmt
* ignore errors when dumping logs and container is missing * fixed typo * print correct payload length * increase conversion rate a bit when estimating fee (to avoid message rejects when rate update tx is active) * fmt
* ignore errors when dumping logs and container is missing * fixed typo * print correct payload length * increase conversion rate a bit when estimating fee (to avoid message rejects when rate update tx is active) * fmt
* ignore errors when dumping logs and container is missing * fixed typo * print correct payload length * increase conversion rate a bit when estimating fee (to avoid message rejects when rate update tx is active) * fmt
* ignore errors when dumping logs and container is missing * fixed typo * print correct payload length * increase conversion rate a bit when estimating fee (to avoid message rejects when rate update tx is active) * fmt
* ignore errors when dumping logs and container is missing * fixed typo * print correct payload length * increase conversion rate a bit when estimating fee (to avoid message rejects when rate update tx is active) * fmt
* ignore errors when dumping logs and container is missing * fixed typo * print correct payload length * increase conversion rate a bit when estimating fee (to avoid message rejects when rate update tx is active) * fmt
* ignore errors when dumping logs and container is missing * fixed typo * print correct payload length * increase conversion rate a bit when estimating fee (to avoid message rejects when rate update tx is active) * fmt
* ignore errors when dumping logs and container is missing * fixed typo * print correct payload length * increase conversion rate a bit when estimating fee (to avoid message rejects when rate update tx is active) * fmt
* ignore errors when dumping logs and container is missing * fixed typo * print correct payload length * increase conversion rate a bit when estimating fee (to avoid message rejects when rate update tx is active) * fmt
* ignore errors when dumping logs and container is missing * fixed typo * print correct payload length * increase conversion rate a bit when estimating fee (to avoid message rejects when rate update tx is active) * fmt
* ignore errors when dumping logs and container is missing * fixed typo * print correct payload length * increase conversion rate a bit when estimating fee (to avoid message rejects when rate update tx is active) * fmt
Fixes #4206 In #1340 one of the storage types was changed from `Candidates` to `CandidateList`. Since the actual key includes the hash of this value, all of the candidates stored here are (a) "missing" and (b) unable to unreserve their candidacy bond. This migration kills the storage values and refunds the deposit held for each candidate. --------- Signed-off-by: georgepisaltu <george.pisaltu@parity.io> Co-authored-by: georgepisaltu <52418509+georgepisaltu@users.noreply.github.com> Co-authored-by: Oliver Tale-Yazdi <oliver.tale-yazdi@parity.io> Co-authored-by: georgepisaltu <george.pisaltu@parity.io>
Fixes #4206 In #1340 one of the storage types was changed from `Candidates` to `CandidateList`. Since the actual key includes the hash of this value, all of the candidates stored here are (a) "missing" and (b) unable to unreserve their candidacy bond. This migration kills the storage values and refunds the deposit held for each candidate. --------- Signed-off-by: georgepisaltu <george.pisaltu@parity.io> Co-authored-by: georgepisaltu <52418509+georgepisaltu@users.noreply.github.com> Co-authored-by: Oliver Tale-Yazdi <oliver.tale-yazdi@parity.io> Co-authored-by: georgepisaltu <george.pisaltu@parity.io>
Fixes #4206 In #1340 one of the storage types was changed from `Candidates` to `CandidateList`. Since the actual key includes the hash of this value, all of the candidates stored here are (a) "missing" and (b) unable to unreserve their candidacy bond. This migration kills the storage values and refunds the deposit held for each candidate. --------- Signed-off-by: georgepisaltu <george.pisaltu@parity.io> Co-authored-by: georgepisaltu <52418509+georgepisaltu@users.noreply.github.com> Co-authored-by: Oliver Tale-Yazdi <oliver.tale-yazdi@parity.io> Co-authored-by: georgepisaltu <george.pisaltu@parity.io>
Fixes #4206 In #1340 one of the storage types was changed from `Candidates` to `CandidateList`. Since the actual key includes the hash of this value, all of the candidates stored here are (a) "missing" and (b) unable to unreserve their candidacy bond. This migration kills the storage values and refunds the deposit held for each candidate. --------- Signed-off-by: georgepisaltu <george.pisaltu@parity.io> Co-authored-by: georgepisaltu <52418509+georgepisaltu@users.noreply.github.com> Co-authored-by: Oliver Tale-Yazdi <oliver.tale-yazdi@parity.io> Co-authored-by: georgepisaltu <george.pisaltu@parity.io>
Fixes #4206 In #1340 one of the storage types was changed from `Candidates` to `CandidateList`. Since the actual key includes the hash of this value, all of the candidates stored here are (a) "missing" and (b) unable to unreserve their candidacy bond. This migration kills the storage values and refunds the deposit held for each candidate. --------- Signed-off-by: georgepisaltu <george.pisaltu@parity.io> Co-authored-by: georgepisaltu <52418509+georgepisaltu@users.noreply.github.com> Co-authored-by: Oliver Tale-Yazdi <oliver.tale-yazdi@parity.io> Co-authored-by: georgepisaltu <george.pisaltu@parity.io>
Fixes #106
Port of cumulus PR paritytech/cumulus#2960
This PR adds the ability to bid for collator slots even after the max number of collators have already registered. This eliminates the first come, first served mechanism that was in place before.
Key changes:
update_bond
extrinsic to allow registered candidates to adjust their bonds in order to dynamically control their bidstake_candidate_slot
extrinsic to try to replace an already existing candidate by bidding more than themDesiredCandidates
out ofMaxCandidates
candidates in the list will be selected by the session pallet as collatorsset_candidacy_bond
call, candidates which don't meet the new bond requirements are kickedChecklist
T
required)