In mid‑January 2025 a developer team noticed a flood of 404 responses coming from BunnyCDN’s storage layer. Files that had been uploaded through the official API disappeared within hours, yet neither the client nor Bunny’s logs recorded a delete operation. Support confirmed the anomaly on Jan 14, saying the objects existed in a replication zone but were absent from the primary region.

The ticket opened on Jan 13 quickly escalated through four different support agents, but each reply merely forwarded the issue to an internal storage team. By Jan 17 the missing‑file count topped 200+ across seven days, and by April 2025 the problem persisted despite repeated bumps. Bunny’s engineers finally admitted the bug in March 2026, attributing it to recent deployment changes meant to improve resilience.

For customers who rely on BunnyCDN for production assets, the silent data loss undermines any confidence in its durability guarantees. The incident illustrates how insufficient visibility into replication zones can mask systemic faults, forcing users into costly migrations they may lack resources for. Reliability reports now list BunnyCDN among high‑risk providers. The reporter’s final verdict: avoid Bunny storage for any critical workload.