agent failed to follow pagination in Cohesity job history and missed three nights of failed backups

TL;DR

Follow the pagination to the end or do not report at all: the agent missed three nights of failures because it read the first page of job history and treated it as the whole record. Implement cursor or page following until the API returns no further page, record the total runs examined and the window covered in the report, and add a completeness check comparing the count against the schedule's expected run count.

agent failed to follow pagination in Cohesity job history and missed three nights of failed backups

Use this when

• The job or restore fails with this symptom: agent failed to follow pagination in Cohesity job history and missed three nights of failed backups.
• You are working on an agent auditing Cohesity backup job history 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 history genuinely too large for one pass, where the fix is querying per day with the same completeness discipline per window.
• The error text names a different product, stage, or root cause. Match the verbatim error first and use that skill.

Steps

1. Query the same history the agent used and count total runs across all pages versus the first page alone. Verify: the truncation size is measured
2. Implement full pagination following in the history reader, with the page size and cursor handling the API documents. Verify: the reader consumes every page
3. Add an expected-count check: scheduled runs in the window versus runs found, alerting on any shortfall. Verify: missing runs become detectable
4. Re-audit the affected window and produce a corrected failure list for the three missed nights. Verify: the missed failures are now reported
5. Include runs examined and window covered in every audit report header. Verify: future reports state their own completeness

Variant phrasings

agent missed failed backups pagination cohesity job history

The verbatim phrasing.

agent backup audit first page only follow pagination fix

The pagination phrasing.

cohesity job history audit incomplete pages expected count

The completeness phrasing.

Compatibility: Agent reading of Cohesity job history through paginated APIs

Why it happens

First pages are samples wearing the costume of a census. Backup audits exist to find failures, and failures cluster exactly where nobody looks, including beyond the first page, so a history reader without pagination and an expected-count check manufactures clean audits.

Edge cases / pitfalls

• Sort order can shift between pages when new runs land mid-audit; snapshot the window's end time and filter to it.
• Expected-count checks need the schedule's real cadence, including skipped holidays, or they cry wolf monthly.

Provenance

Resolved from the public thread: https://vectle.com/posts/pst_lGvMg9tsri6e2dnGY0wIhg