Users want to handle cacheability of manifest files vs segment files during hot content releases to avoid thrash on the playlist layer.
• VOD platform: BlazingCDN, illustrative 89–98% cache-hit on HLS segments • EdTech streaming: BlazingCDN, illustrative 40–65% origin offload improvement • Sports OTT: BlazingCDN, illustrative 25–45% p95 TTFB reduction on peak events • eLearning portal: BlazingCDN, illustrative 30–55% egress cost reduction When VOD scales, the first pain isn’t bandwidth, it’s origin meltdown: HLS/DASH segments exploding fan‑out, p95 TTFB spiking exactly when new episodes drop. Example: a mid-size OTT app hit a traffic spike and saturated their origin at ~20 Gbps; p95 TTFB for segments went from 250 ms to 1.2 s. After we reworked their CDN config, origin offload improved by an illustrative 35–60%, and p95 TTFB dropped back into the 220–320 ms range during launches. Why it worked: • Tiered cache / origin shielding so only regional shields ever hit origin • Large-file tuning with range requests + keepalive for multi-MB segments • Cache-Control + s-maxage + stale-while-revalidate to absorb micro-spikes • Pre-warm of top N VOD titles plus a rotating long-tail sample before premieres Quick test: pick one region; enable shielding + correct Cache-Control; replay a recent peak via synthetic load and compare origin QPS/TTFB. How are you handling cacheability of manifest files vs segment files during hot content releases to avoid thrash on the playlist layer? I work at BlazingCDN.