ioGuardian: Real-Time Recovery When Drives and Servers Fail
ioGuardian moves beyond instant recovery to real-time recovery. When failures exceed what synchronous replication can absorb, VergeFS pulls the missing data segments from the ioGuardian server and feeds them to running VMs on the production cluster. Nothing restarts, nothing moves and the IT team does not need to intervene.
When Failures Exceed Protection, Recovery Usually Means a Restore
Synchronous replication handles the common case: RF2 survives one drive or node failure and RF3 survives two. When a failure goes beyond that, most platforms hand the problem to backup software. The team identifies the affected VMs, restores them from a copy typically four to eight hours old and brings them back into service by hand.
Multiple Drive Failures
A second or third drive failure during a rebuild no longer turns into a restore job.
Multiple Host Failures
Losing more servers than the replication level allows does not take VMs offline.
Complex Recovery
No hunting for affected VMs, no restore queue and no manual return to production.
Inline Recovery Keeps VMs on the Production Cluster
ioGuardian is an independent protection layer: one extra server that holds a complete, deduplicated copy of the VergeOS instance. It protects the entire instance, not individual disks or storage groups.
When a requested block cannot be found on the primary storage path, VergeFS checks the local replica first, then the ioGuardian server, then the DR site. The VM keeps running on the primary instance at production performance. The only cost is some added latency during retrieval.
- Near-continuous updates. Send data to the ioGuardian server every hour, every 30 minutes or every 15 minutes.
- Only unique data. Global inline deduplication sends only unique segments and keeps both production and ioGuardian capacity low.
- Flexible media. Asynchronous transfer lets the ioGuardian server use hard drives, even when production is all flash.
- On-site or off-site. The ioGuardian server can sit in another building or another city and serve as your off-site DR copy. Inline delivery still works, with latency as the only trade-off.
Multi-layer retrieval keeps data flowing through complex failures.
Back to Full Protection Without a Restore
When a failed drive is replaced, data streams directly from the ioGuardian server to rebuild it. The cluster returns to a fully protected state with minimal impact on production performance, and no one has to restore volumes by hand.
- Automatic rebuild onto new drives after a failure
- Production VMs keep running throughout
- Works the same for one drive or several
Replacement drives fill from ioGuardian in the background.
The Right Recovery Path for Each Failure
VergeOS snapshots and ioGuardian share the same code base and the same deduplication engine, but they cover different failure domains. Each layer does a job the others leave alone.
| Failure | Recovery Path | What Happens |
|---|---|---|
| Deleted file, corrupted database, ransomware | VergeOS snapshots | Roll back on local deduplicated storage in seconds. Drives are healthy, nothing crosses the network. |
| One or two drive or node failures | RF2 / RF3 replication | Reads redirect to surviving copies, VMs stay online and protection rebuilds in the background. |
| Failures beyond the replication level | ioGuardian | Missing segments delivered inline to running VMs, replacement drives rebuilt from the ioGuardian server. |
| Site outage | ioReplicate | Recover whole Virtual Data Centers at a second site in three clicks. |
| Long-term retention, off-platform restore | Veeam Backup & Replication | Natively supported in VBR 13.1 for retention, immutable offsite copies and application-item recovery. |
One Server. No Extra Software. Off-Site if You Want It.
ioGuardian is built into VergeFS. The only addition is one dedicated server configured and licensed as the ioGuardian target. You license the server, not the feature.
- Use older hardware. The target can be an older server or storage system capable of running VergeOS.
- Minimal capacity planning. Global inline deduplication shrinks the data before it reaches the target.
- Your off-site DR copy. Place the ioGuardian server at a remote site and it doubles as your off-site copy of the data. VMs still receive missing segments inline across the WAN, and the only cost is added latency.
- Or place two. Put one ioGuardian server on-premises for the fastest inline delivery and one at a remote location for disaster protection.
Snapshots Plus ioGuardian
ioGuardian works with VergeOS snapshots, usually taken hourly, to tighten both RPO and RTO. Snapshots can be retained as long as needed without affecting production performance, and more frequent snapshots narrow the RPO further.
Resources
Exit VMware to Resilient HCI
Watch drive and server failures handled live, from a pulled drive to a full failover.
BlogRefurbished Hardware Made Resilient
Why ioGuardian makes used drives and servers a practical way to fund the VMware exit.
On-Demand WebinarHCI and Hard Drives
How hard drives and VergeOS fight data center inflation.
Frequently Asked Questions About ioGuardian
What is ioGuardian?
ioGuardian is an independent protection layer built into VergeFS. One extra server holds a complete, deduplicated copy of the VergeOS instance and delivers missing data segments to running VMs in real time when failures exceed what synchronous replication can absorb.
How is real-time recovery different from instant recovery?
Instant recovery restarts VMs from a backup copy, often on backup storage. ioGuardian keeps VMs running on the production cluster and feeds them the missing data inline, so there is no restart, no failover and no manual return to production.
What failures does ioGuardian protect against?
ioGuardian protects against multiple drive failures and multiple host failures that go beyond the RF2 or RF3 replication level, the cases that would otherwise send recovery back to a restore from backup.
How often is data sent to the ioGuardian server?
Data can be sent almost continuously, every hour, every 30 minutes or every 15 minutes, which keeps the recovery point under an hour. Only unique, deduplicated segments are transferred.
What happens when I replace a failed drive?
Data streams directly from the ioGuardian server to rebuild the new drive. The cluster returns to a fully protected state with minimal impact on production performance and no manual volume restore.
Does ioGuardian cost extra?
The ioGuardian capability is built into VergeFS at no additional cost. Customers license the additional server that acts as the ioGuardian target, not the feature.
What hardware does the ioGuardian server need?
The target can be an older server or storage system capable of running VergeOS. Asynchronous transfer lets it use hard drives, even when production is all flash, and deduplication keeps capacity planning minimal.
How do VergeOS snapshots and ioGuardian work together?
Snapshots handle logical failures such as deleted files, corruption or ransomware, rolling back in seconds when the drives are healthy. ioGuardian handles physical media failures beyond the replication level. Both keep production running on the primary cluster.
Can ioGuardian be my off-site DR copy?
Yes. Place the ioGuardian server at a remote site and it serves as your off-site copy of the data. VMs still receive missing segments inline across the WAN, and the only cost is added latency. For the fastest inline delivery plus disaster protection, place one ioGuardian server on-premises and one off-site.
Do I still need backup software with ioGuardian?
ioGuardian is not a backup appliance. It keeps production running through hardware failures. Backup software such as Veeam Backup & Replication 13.1, which supports VergeOS natively, handles long-term retention, independent offsite copies and restores onto other platforms.
Keep VMs Running Through the Failures That Usually Mean a Restore
Add one server, keep production on the primary cluster and rebuild replacement drives automatically.