Overview
The following graphs show a summary of the output from each of the six steps. Each graph contains a description and bullet points of what to look for.
Data Volume Overview for the lifetime of the system:
- Capacity never flattens out
- Used disk space is currently 75 TB which is 7 TB over the truncation start of 68 TB
- GC is constantly running
- Reduction ratio is 10.91 : 1
- Blockpool is 56.996 TB
Timeline determined from lifetime graph starting at the end of the last major capacity decline:
- Same points as the graph from the lifetime of the system
Space reclamation graph using the timeline
- Stage 4 hasn’t run since Week 52
- Week 05 shows a Stage 2 operation running for over a week then goes to stage 1
- The system is not able to complete GC
Ingest during timeline
- The system is experiencing a lot of Resets and Aborts
- Weeks 04 and 05 show some very large ingest and read patterns
Daily Ingest during timeline
- Daily ingest rates are often twice the recommended maximum of 6 TB per day and are 3 times as much for a couple of days in weeks 04 and 05.
Capacity - Data Volume Overview
- Capacity never flattens out
- GC is constantly running
- Reduction ratio is 10.98 : 1
- Blockpool is 58.373 TB
- There is an increase in blockpool capacity of 5 TB for the previous 9 weeks
Capacity - Used Disk Space during timeline
- Spikes above the Truncation Start line indicate dedup backlog
- There was a serious capacity problem between weeks 05 and 07
Reduction Ratio during timeline:
- Reduction ratio is approximately 10 which is acceptable
Disk I/O during timeline
- Disk I/O is under or at 70% which is acceptable
- Over 70% is not acceptable
Here are all of the analysis points without screenshots:
Data Volume Overview for the lifetime of the system
- Capacity never flattens out
- Used disk space is currently 75 TB which is 7 TB over the truncation start of 68 TB
- GC is constantly running
- Reduction ratio is 10.91 : 1
- Blockpool is 56.996 TB
Timeline determined from lifetime graph starting at the end of the last major capacity decline:
- Same points as the graph from the lifetime of the system
Space reclamation graph using the timeline
- Stage 4 hasn’t run since Week 52
- Week 05 shows a Stage 2 operation running for over a week then goes to stage 1
- The system is not able to complete GC
Ingest during timeline
- The system is experiencing a lot of Resets and Aborts
- Weeks 04 and 05 show some very large ingest and read patterns
Daily Ingest during timeline
- Daily ingest rates are often twice the recommended maximum of 6 TB per day and are 3 times as much for a couple of days in weeks 04 and 05.
Capacity - Data Volume Overview
- Capacity never flattens out
- GC is constantly running
- Reduction ratio is 10.98 : 1
- Blockpool is 58.373 TB
- There is an increase in blockpool capacity of 5 TB for the previous 9 weeks
Capacity - Used Disk Space during timeline
- Spikes above the Truncation Start line indicate dedup backlog
- There was a serious capacity problem between weeks 05 and 07
Reduction Ratio during timeline:
- Reduction ratio is approximately 10 which is acceptable
Disk I/O during timeline
- Disk I/O is under or at 70% which is acceptable
- Over 70% is not acceptable
Conclusion
After using the methods in this playbook we can say, with confidence, that the customer is ingesting too much data.
Positive performance points:
- Reduction ratio looks fine
- Disk I/O is OK
- Blockpool is at 65% of total capacity
Negative performance points:
- Daily ingest is too high
- GC can’t complete and won’t catch up
- Dedup is sometimes backlogged for weeks
Resolution options:
- Implement additional DXi systems
- Reduce the amount of data that is being ingested
Quantum