Remove failing assertion related to VoterList count mismatch #10880

cirko33 · 2026-01-22T14:54:31Z

PR resolves the error that occurs when VoterList count does not align with sum of Nominators & Validators in bags-list. This misalignment happens while bags-list is locked & nominations/validations are received. Eventually this misalignment evens out due PendingRebag implementation

cirko33 · 2026-01-22T15:07:14Z

/cmd prdoc --audience runtime_dev --bump patch

…time_dev --bump patch'

sigurpol · 2026-01-22T15:57:14Z

prdoc/pr_10880.prdoc

+doc:
+- audience: Runtime Dev
+  description: |-
+    Refers to: #10687


I know the bot has generated this prdoc starting from the description you put in the PR but usually we prefer not to link explicitly to other GH issues / PRs. Just a very syntethic description of the problem you are trying to fix would be better

sigurpol · 2026-01-22T16:02:30Z

substrate/frame/staking-async/src/pallet/mod.rs

-				"not all genesis stakers were inserted into sorted list provider, something is wrong."
-			);
+			// Due to `PendingRebag`, the bags-list may temporarily lag behind the true
+			// voter set in `pallet-staking`, making such an assertion invalid.


Rather than commenting out, I would evaluate if we might see value in changing the assertion into just logging for debugging purposes (leaving the comment about why this is not a valid invariant).

Pedantic: technically this PR could be backported to a runtime that doesn't have PendingRebag implementation (#9725, currently on 2509 / 2512 but not on 2507). Probably we won't make a new 2507 release backporting this PR and not #9725 but maybe we can rephrase better.

Also it's not because PendingRebag that we lag behind but it's because PendingRebag that eventually we converge and get rid of misalignment

cirko33 · 2026-01-22T18:29:07Z

/cmd prdoc --audience runtime_dev --bump patch

github-actions · 2026-01-22T18:31:50Z

Command "prdoc --audience runtime_dev --bump patch" has failed ❌! See logs here

cirko33 · 2026-01-22T18:37:32Z

/cmd prdoc --audience runtime_dev --bump patch --force true

…time_dev --bump patch --force true'

sigurpol · 2026-01-23T08:56:26Z

Good job! I would suggest to add an integration test under ahm-tests that simulates the following scenario reported by @Ank4n in #10687 .

- While election snapshot is ongoing, voterlist is locked, and no updates happen to it.
- If a validator starts validating, or a nominator nominates for the first time during these snapshot blocks, the account never ends up in the bags list, and therefore never gets exposed.

The test would need to verify that, in such a case, the PendingRebag mechanism ensures that once the voterlist is unlocked, validators/nominators are successfully inserted into the bags lists.

We have unit tests in bags-list pallet such as pending_rebag and rebag_extrinsic_removes_from_pending testing in isolation that lock bag - try to insert - add to pending rebag - unlock - on_idle or explicit extrinsic rebag sanitizes bag list.

I believe we might not have a more comprehensive integration test (to be checked!!!) so that would be a good excuse for you to look at integration tests in staking as well 😉

sigurpol · 2026-01-23T09:01:03Z

prdoc/pr_10880.prdoc

+title: Remove failing assertion related to VoterList count mismatch
+doc:
+- audience: Runtime Dev
+  description: PR resolves the error that occurs when VoterList count does not align


small tip or at least my preference: Once you generate the prdoc the first time with /cmd prdoc --audience runtime_dev --bump patch , I usually prefer not to re-enter /cmd prdoc --audience runtime_dev --bump patch --force true if I need to modify prdoc further, I'll just edit manually in my editor and push together with my other changes (pay attention to indentation, there are some guides in the how-to-contribute guide or similar, and some plugin for some popular Microsfot editor I refuse to use ;-) ). This allows you to easily decouple what you have in the PR message that can be more verbose and what you have in the prdoc that must be kept to the minimum. In the PR message for example it would be valuable to mention Close #10687 so that it automatically closes that issue once we merge. And you can provide more details or explanation if you feel they are needed

sigurpol · 2026-01-23T09:09:30Z

prdoc/pr_10880.prdoc

+doc:
+- audience: Runtime Dev
+  description: PR resolves the error that occurs when VoterList count does not align
+    with sum of Nominations & Validators in bags-list. This misalignment happens while


[nit] this might be improved probably since VoterList is associated to bags-list, not Nominators and Validators. Also Nominations -> Nominators.

Something like this maybe?

The invariant in try-state related to VoterList in bags-list matching the sum of Nominations and Validators no longer holds. The misalignment arises when the bags-list is locked (e.g. during election Snapshot phase) and nominators / validators try to nominate / validate. This discrepancy is eventually reconciled via the PendingRebag mechanism (automatically via on_idle or explicitly via rebag() extrinsic)

kianenigma · 2026-01-23T09:30:57Z

substrate/frame/staking-async/src/pallet/mod.rs

-				Nominators::<T>::count() + Validators::<T>::count(),
-				"not all genesis stakers were inserted into sorted list provider, something is wrong."
+				Nominators::<T>::count(),
+				Validators::<T>::count()


The actual failing assertion was in fn check_count() though, right?

yes

[2026-01-21T23:35:44Z ERROR runtime::frame-support] ❌ "Staking" try_state checks failed: Other("wrong external count")

@cirko33 look at https://github.com/paritytech/polkadot-sdk/blob/master/substrate/frame/staking-async/src/pallet/impls.rs#L1904-L1904

so the current code here should stay as it was, and this needs to be modified

ensure!( <T as Config>::VoterList::count() == Nominators::<T>::count() + Validators::<T>::count(), "wrong external count" );

updates: - revert back to first assertion in `mod.rs` - log instead of assert in `check_count()` function - update prdoc

paritytech-workflow-stopper · 2026-01-23T16:35:15Z

All GitHub workflows were cancelled due to failure one of the required jobs.
Failed workflow url: https://github.com/paritytech/polkadot-sdk/actions/runs/21292677879
Failed job name: test-linux-stable

remove invalid voter count assertion due to PendingRebag

d51249c

cirko33 self-assigned this Jan 22, 2026

cirko33 requested a review from a team as a code owner January 22, 2026 14:54

cirko33 added the A4-backport-stable2512 Pull request must be backported to the stable2512 release branch label Jan 22, 2026

Merge branch 'master' into cirko33-10687-bug

5069946

Update from github-actions[bot] running command 'prdoc --audience run…

17d1d36

…time_dev --bump patch'

sigurpol requested review from Ank4n, andreitrand, kianenigma and sigurpol and removed request for a team January 22, 2026 15:37

sigurpol reviewed Jan 22, 2026

View reviewed changes

cirko33 changed the title ~~remove invalid voter count assertion due to PendingRebag~~ Remove failing assertion related to VoterList count mismatch Jan 22, 2026

log instead of commenting out

9b4583b

Update from github-actions[bot] running command 'prdoc --audience run…

21e3c1c

…time_dev --bump patch --force true'

sigurpol reviewed Jan 23, 2026

View reviewed changes

kianenigma reviewed Jan 23, 2026

View reviewed changes

log instead of assert

9f90bb0

updates: - revert back to first assertion in `mod.rs` - log instead of assert in `check_count()` function - update prdoc

Remove failing assertion related to VoterList count mismatch #10880

Are you sure you want to change the base?

Remove failing assertion related to VoterList count mismatch #10880

Uh oh!

Conversation

cirko33 commented Jan 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

cirko33 commented Jan 22, 2026

Uh oh!

sigurpol Jan 22, 2026

Choose a reason for hiding this comment

Uh oh!

sigurpol Jan 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

cirko33 commented Jan 22, 2026

Uh oh!

github-actions bot commented Jan 22, 2026

Uh oh!

cirko33 commented Jan 22, 2026

Uh oh!

sigurpol commented Jan 23, 2026

Uh oh!

sigurpol Jan 23, 2026

Choose a reason for hiding this comment

Uh oh!

sigurpol Jan 23, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

kianenigma Jan 23, 2026

Choose a reason for hiding this comment

Uh oh!

sigurpol Jan 23, 2026

Choose a reason for hiding this comment

Uh oh!

sigurpol Jan 23, 2026

Choose a reason for hiding this comment

Uh oh!

sigurpol Jan 23, 2026

Choose a reason for hiding this comment

Uh oh!

paritytech-workflow-stopper bot commented Jan 23, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

cirko33 commented Jan 22, 2026 •

edited

Loading

sigurpol Jan 22, 2026 •

edited

Loading

sigurpol Jan 23, 2026 •

edited

Loading