Polkassembly Logo

Head 1
Head 3
Head 4
Create Pencil IconCreate
TRACKS
ORIGINS
Report an issueNeed help with something?
Foot 1
Foot 2
Foot 3
Foot 4
OpenGov
View All Staking Admin
Execution Failed

Cancel Deferred Slash 7126 - Due to application bug

inStaking Admin
10 months ago

Dear community,

I'm submitting this proposal to cancel a "deferred" slash for era 7126 which would be applied in era 7154.  The issue resulting in the slash was deemed to be bug related and the technical details are still being investigated.  Historically, deferred slashes due to bugs were cancelled.

The issue, in my simple interpretation is as follows:

  • There was downtime due to hardware issues
  • Before the downtime one of my nodes was due to issue a "pre-vote"
  • When the node was online again, time passed and the prevote was no longer necessary.
  • The node may have used the local (not sync'd) finalized block as reference and proceeded to issue the pre-vote
  • Because the pre-vote already issued it resulted in an equivocation which is a slashable offence
  • Ideally, the node should have determined that time passed and the pre-vote wasn't necessary anymore

Given the circumstances and if my interpretation is correct this can happen to any operator who may restart their nodes with a pre-vote "in queue" given a delay in finalized blocks.

The issue was reported immediately but because of the bug, the "slash" was reported some 24 later. Time was also lost while trying to get a technical verdict on whether this was a bug or operator related. Given these delays, it is unlikely to receive a cancellation of the deferral on time. I would however engage the community for a deferral in solidarity.

I have already compensated the slashed parties (rounding up) and I am working on another payout for those affected by any offline time.

Kind Regards,
Will | Paradox

Comments (1)

9 months ago

why did operator not chill the validator when there clearly was 2 session downtime after unclean shutdown?

PleaseLogin to comment

Proposal Failed

3

of 3

Summary

0%

Aye

AyeNay

0%

Nay

Aye (62)0.0 DOT

Support0.0 DOT

Nay (3)0.0 DOT

Voting Data

Approval%

Support%

Threshold0.00%

Threshold0.00%

Comments (1)

9 months ago

why did operator not chill the validator when there clearly was 2 session downtime after unclean shutdown?

PleaseLogin to comment

Help Center

Report an Issue
Feedback
Terms and Conditions
Github

Our Services

Docs
Terms of Website
Privacy Policy

A House of Commons Initiative.

Polka Labs Private Limited 2025

All rights reserved.

Terms and ConditionsTerms of Website
Privacy Policy