Commvault is incredibly powerful.
But that power comes with complexity—and complexity is where mistakes creep in.
I’ve worked in enough environments to see the same patterns repeat:
- Backups are “working”… until they aren’t
- Performance slowly degrades
- Restores become harder than they should be
And most of the time, it’s not a product issue.
It’s configuration.
Here are 10 Commvault configuration mistakes I see all the time—and how to avoid them.
1. Building Everything Around a Single MediaAgent
It works at first.
One MediaAgent. One place for everything. Simple.
Until:
- Jobs stack up
- Throughput drops
- That one system becomes your bottleneck
Fix:
Start with at least two MediaAgents and distribute workloads.
If everything depends on one MediaAgent, you don’t have resilience—you have risk.
2. Undersizing the CommServe
The CommServe is the brain of your environment.
And yet, it’s often treated like an afterthought.
What happens:
- Slow job scheduling
- UI lag
- Reporting delays
Fix:
- Proper CPU/RAM sizing
- Fast storage for the database
- Regular DB maintenance
If the CommServe struggles, everything struggles.
3. Poor Deduplication Database (DDB) Placement
This one causes more long-term pain than almost anything else.
The mistake:
Putting the DDB on slow or shared storage.
The result:
- Slower backups
- Longer job times
- Painful rebuilds
Fix:
- Place DDB on fast, dedicated storage (SSD if possible)
- Monitor DDB health regularly
Dedupe performance lives and dies by storage speed.
4. Overloading a Single Storage Pool
It’s easy to just keep adding workloads to the same storage pool.
Until it can’t keep up.
Symptoms:
- Increased job duration
- Resource contention
- Inconsistent performance
Fix:
- Split workloads across multiple storage pools
- Align pools with workload types or performance tiers
One pool for everything = one problem for everything.
5. Ignoring Network Throughput
Backup traffic is still network traffic.
And it’s often underestimated.
What happens:
- Bottlenecks between clients and MediaAgents
- Slower backups and restores
- Unpredictable performance
Fix:
- Validate bandwidth between components
- Use dedicated backup networks where possible
- Monitor throughput, not just job status
You can’t out-configure a slow network.
6. No Clear Retention Strategy
Retention tends to evolve organically—and that’s the problem.
What I see:
- Different retention policies everywhere
- No alignment with business needs
- Storage filling up unexpectedly
Fix:
- Define retention by workload type
- Align with compliance and business requirements
- Plan for growth, not just current usage
Retention isn’t a setting—it’s a strategy.
7. Mixing All Workloads Together
Databases. VMs. File servers. Archive jobs.
All in the same policies, same schedules, same infrastructure.
What happens:
- Performance conflicts
- Harder troubleshooting
- Unpredictable job behavior
Fix:
Segment workloads:
- By type
- By priority
- By performance requirements
Separation creates control—and control creates stability.
8. Skipping Restore Testing
This is the big one.
Everything looks fine… because nobody has tried to restore anything.
Reality:
- Backups can succeed but still be unusable
- Application restores may fail due to misconfigurations
- Recovery times are often unknown
Fix:
Test regularly:
- Full VM restores
- File-level recovery
- Application-level recovery
Backup success means nothing without recovery success.
9. Too Many Exceptions (No Standardization)
“I’ll just configure this one differently…”
That adds up quickly.
What I see:
- Inconsistent policies
- Confusing configurations
- Hard-to-manage environments
Fix:
- Standardize naming, policies, and schedules
- Limit exceptions
- Document anything that deviates
Consistency is what makes environments scalable.
10. Ignoring Alerts and Warning States
Commvault will tell you when something isn’t right.
The problem is… people stop listening.
What happens:
- Warnings pile up
- Real issues get buried
- Small problems become big ones
Fix:
- Clean up existing warnings
- Tune alerts so they’re meaningful
- Treat warnings as early signals—not noise
If everything is a warning, nothing is.
Bringing It All Together
Most Commvault issues don’t come from dramatic failures.
They come from:
- Small misconfigurations
- Overlooked details
- Decisions that made sense at the time
Until scale exposes them.
Final Thought
You don’t need a perfect environment.
But you do need a predictable one.
Good configuration isn’t about making things work today.
It’s about making sure they still work when everything grows.
Fix these early, and your environment will feel stable.
Ignore them, and you’ll spend your time chasing problems that didn’t have to exist.
