Our MongoDB replica set is experiencing a storage imbalance across nodes. We have identified that one secondary node is consuming significantly more disk space compared to the primary and other secondary nodes, which suggests potential data fragmentation or oplog sizing issues.

Potential Causes to Investigate:

1. Oplog Fragmentation: Secondary node may have excessive oplog fragmentation
2. Index Fragmentation: Indexes on secondary might be fragmented
3. Delayed Replication: Node may be behind in replication causing log buildup
4. WiredTiger Storage Engine Issues: Collection/index fragmentation
5. Oplog Size Configuration: Mismatched oplog size settings across nodes

Troubleshooting Attempted

• Compact command execution - No effect
• Oplog size window modification - No effect