Zerto VPG Monitoring: Service Overview
Overview
We monitor your Zerto Virtual Protection Groups (VPGs) for issues that put your ability to recover at risk, and we act on the ones that persist. Our monitoring is designed to surface real, sustained replication problems while screening out routine activity such as DR tests, planned changes and pauses you start yourself.
This document explains how the monitoring works, which conditions generate an alert to our team, and which conditions are intentionally excluded.
How monitoring works
An alert reaches our team only after an issue has been detected on six consecutive checks, with checks running every six hours. This confirms the problem is sustained and not a short-lived or planned event.
- Collect. Zerto status and alert data for each of your VPGs is gathered continuously into our monitoring platform.
- Check. Every six hours, an automated rule evaluates each VPG against the criteria in the next two sections.
- Confirm. A VPG that meets the alert criteria is flagged. It must be flagged on six consecutive checks before an alert is raised. If it recovers in between, the count resets.
- Alert. Confirmed issues are automatically passed to our operations team for review.
- Respond. Our team investigates, remediates where it is within our scope, and contacts you when your input or action is needed.
What is monitored
We raise an alert when a production VPG shows a sustained error state or a recovery point that is broken or more than six hours old.
Condition | What it means for you | Alert threshold |
|---|---|---|
VPG health in an error state | Replication for the VPG is not working as expected | Any status other than Healthy or Warned |
Recovery point objective (RPO) exceeded | Your most recent recoverable point is more than 6 hours old | RPO above 21,600 seconds (6 hours) |
Broken RPO | The VPG has no valid recovery point and is not protecting your data | RPO reported below 0 seconds |
Other Zerto alerts | Zerto has raised an alert not listed under exclusions below | Any alert type not excluded |
All conditions must persist across six consecutive checks before an alert is raised, as described above.
What is not monitored
The following are excluded from alerting because they reflect normal, planned or client-controlled activity, or because they relate to configuration and capacity rather than a replication failure.
Excluded VPGs and states
Exclusion | Why it is excluded |
|---|---|
VPGs with "test" or "migration" in the name | These are treated as non-production and temporary |
VPG health of Healthy or Warned | These statuses do not indicate a failure |
RPO between 0 and 6 hours | Covers normal replication lag and planned activity such as DR tests |
VPGs paused by a user | A pause you initiate is treated as intentional |
Issues that clear within six consecutive checks | Short-lived and self-resolving issues are not escalated |
Excluded Zerto alert types
Zerto alert | Description |
|---|---|
VPG0003 / VPG0004 | Journal history is lower than configured |
VPG0011 | The scheduled test interval passed without the VPG being tested |
VPG0035 / VPG0036 | VPG protected resources are not assigned to the Zerto Organization (ZORG) |
VPG0037 / VPG0038 | Configured journal history is longer than the history currently available |
VPG0063 | The Zerto Organization (ZORG) does not exist |
VRA0030 | A previous journal configuration update for a VM in the VPG failed |
VRA0039 | Virtual Replication Appliance (VRA) reconciliation |
VRA0040 | A journal is nearly full with limited space remaining |
VRA0060 | A malformed packet was received on the host |
If one of these conditions leads to a broken or overdue RPO or an error health state, that downstream impact is still detected and alerted as described above.