Details
-
Untriaged
-
-
0
-
Unknown
-
Analytics Sprint 46
Description
Steps we did:
1. Create a 2 node 8vCPU, 32GiB and a capella provisioned cluster with 2 buckets |
2. Create 5 remote links, 50 remote collection, 1 S3 link and 1 S3 collection. |
3. Start data load on provisioned with 20GB per bucket - Total 1TB in columnar and scale the columnar instance to 8 nodes |
4. Wait for ingestion complete and take a backup of the columnar instance |
5. Restore the backup |
6. Start data load on provisioned to reach 40GB per bucket - Total 2TB in columnar |
7. Wait for load to complete and take the backup |
8. Start data load on provisioned with 60GB per bucket - Total 3TB in columnar |
9. Wait for ingestion complete and restore with 2TB backup |
10. Wait for ingestion complete (columnar have 3TB after this step) |
11. Start a backup job |
12. While backup is in progress scale down the instance to 2 nodes. |
The cluster becomes unusable for 50 minutes after the scaling is complete |
It can be because of the cc node get rebalanced out
References: https://issues.couchbase.com/browse/MB-62680?focusedId=785211&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-785211
CC Node Ticket: https://couchbasecloud.atlassian.net/browse/AV-82021
Attachments
Issue Links
- duplicates
-
MB-62680 [System Test] [Intermittent] Cluster unusable for a brief period after scale down operations
- Open