agent conflated snapshot presence with backup success for Zerto VPGs in a replication lag report

TL;DR

Snapshot or checkpoint presence is not replication health for Zerto: health is the VPG's RPO against its SLA and its sync state, and a VPG can hold checkpoints while lagging hours behind. Report per VPG the current RPO, the target RPO, and the state, flag any VPG whose lag exceeds its target regardless of how many checkpoints exist, and trend the lag so a slowly drifting VPG is caught before it breaches.

agent conflated snapshot presence with backup success for Zerto VPGs in a replication lag report

Use this when

• The job or restore fails with this symptom: agent conflated snapshot presence with backup success for Zerto VPGs in a replication lag report.
• You are working on an agent reporting Zerto replication lag for protected VPGs and the failure repeats on the same target or schedule.
• You need the failing job, stage, or component isolated before changing retention or infrastructure.

Not for this skill when

• A VPG in a planned initial sync whose lag is expected and scheduled, which should be reported as in-progress with its completion estimate, not as a breach.
• The error text names a different product, stage, or root cause. Match the verbatim error first and use that skill.

Steps

1. Pull each VPG's current RPO, target RPO, and state from Zerto, not its checkpoint count. Verify: the health inputs are the right ones
2. Recompute the report's health per VPG as lag versus target, listing breaches by size. Verify: the corrected health picture is produced
3. Compare with the previous report to find any breach the snapshot-based version masked. Verify: masked breaches are surfaced
4. Rewrite the report generator to use RPO and state, with checkpoint counts removed or demoted to context. Verify: the generator measures replication, not artefacts
5. Add lag trending with an alert at a fraction of the target so drift pages before breach. Verify: early warning exists for the next drift

Variant phrasings

agent conflated snapshot presence backup success zerto vpg lag

The verbatim phrasing.

zerto replication health rpo not checkpoints agent report fix

The RPO phrasing.

agent zerto lag report wrong metric vpg state rpo

The metric phrasing.

Compatibility: Agent reporting of Zerto replication health from VPG state

Why it happens

Continuous replication succeeds or fails as a rate, data copied versus data changing, and artefacts accumulate on both sides of that race. Counting artefacts to judge a rate is a category error that reports green precisely while the gap the report exists to expose grows.

Edge cases / pitfalls

• During initial sync, lag is structural; the report should show progress and estimated completion rather than a binary state.
• A VPG meeting RPO while in a warning state still deserves a line; state and lag are separate axes.

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_l78IQZqsX2bSVC-ZxLbADw