Bound some storage items for pallet staking and clean up deprecated exposures
#7483
Conversation
re-gius
changed the title
staking and clean up deprecated exposuresstaking and clean up deprecated exposures
|
/cmd prdoc --audience runtime_dev --bump patch |
|
Deferring |
| @@ -842,7 +802,11 @@ pub mod pallet { | |||
| fn build(&self) { | |||
| ValidatorCount::<T>::put(self.validator_count); | |||
| MinimumValidatorCount::<T>::put(self.minimum_validator_count); | |||
| Invulnerables::<T>::put(&self.invulnerables); | |||
| assert!( | |||
| self.invulnerables.len() as u32 <= T::MaxInvulnerables::get(), | |||
There was a problem hiding this comment.
self.invulnerables is already bounded, we don't need this check.
| let disabled_validators_maybe = BoundedVec::try_from(old_disabled_validators); | ||
| match disabled_validators_maybe { | ||
| Ok(disabled_validators) => { | ||
| y = y.saturating_add(1); |
There was a problem hiding this comment.
You have read the storage already no matter if the match is Ok or Err.
| match disabled_validators_maybe { | ||
| Ok(disabled_validators) => { | ||
| y = y.saturating_add(1); | ||
| DisabledValidators::<T>::set(disabled_validators); |
There was a problem hiding this comment.
The main point that might be missed here is that the encoding of a Vec<_> and BoundedVec<_> is exactly the same. So no data actually needs to be migrated.
What the migration would have to do is to check if the data won't fit into BoundedVec<_>, and in this case truncate it.
If hypothetically DisabledValidators is too large, your migration will not touch it, which means the runtime will still fail to decode it upon next access.
| log!(warn, "v17 failed to bound some storage items."); | ||
| } | ||
|
|
||
| T::DbWeight::get().reads_writes(x.into(), y.into()) |
There was a problem hiding this comment.
In any case I will temp remove this migration as we will start from a fresh state and migrate the data. So any failed bounding will be detected in the migration testing.
| <Pallet<T>>::deposit_event(super::Event::<T>::ValidatorDisabled { | ||
| stash: params.stash.clone(), | ||
| }); | ||
| if disabled.try_insert(index, (offender_idx, new_severity)).is_ok() { |
There was a problem hiding this comment.
For this you can use .defensive, worth reading up on it.
|
All GitHub workflows were cancelled due to failure one of the required jobs. |
Building from #6445 on top of #7282
Changes
Invulnerables, vector of validators invulnerable to slashing.MaxInvulnerablesto boundInvulnerablesVec ->BoundedVec.westend).Disabled Validators, vector of validators that have offended in a given era and have been disabled.MaxDisabledValidatorsto boundDisabledValidatorsVec ->BoundedVec.MaxValidatorsCountaccording to the current disabling strategy).ErasStakersandErasStakersClipped(see Tracker issue for cleaning up old non-paged exposure logic in staking pallet #433 ), non-paged validators exposures.Migrating pallet
stakingstorage to v17 to apply all changes.TO DO (in a follow-up PR)
ErasStakersPagedExposurePage.othersvectorBondedErasvectorClaimedRewardspages vectorErasRewardPointsEraRewardPoints.individualBTreeMapUnappliedSlashesSlashingSpansSlashingSpans.priorvector