iops
Proactive Warnings for Concurrent Operations
The Problem: We experienced 1+ hour service outage because we made two separate changes (IOPS ) too close together.
We can see the guidance in docs but it cannot proactively prevent the issue.
from 3000 to 6000 : There was no node down issue
from 6000 to 8000 : Issue happened.
The Feature Request: Add a proactive warning system that detects when a cluster modification is already in progress. If a user tries to submit a second change, an alert should pop up explaining that a prior operation is still underway and that submitting a second change could result in an outage. This prevents accidental, concurrent changes from the start.
Seokwoo
shared this idea
-
soyee
commented
We need the same feature request since we experienced the similar issue as below.
We experienced 2+ hours service outage because we made two separate changes (IOPS and Storage) too close together.
We can see the guidance in docs but it cannot proactively prevent the issue.
• On 06/16/25 - 06:04:22 AM UTC customer updated the IOPS from 3000 to 1000 IOPS.
• On 06/16/25 - 07:13:01 AM UTC customer scaled the volume from 600GB to 1200GB.