Purge All

​

Use this when

• You can’t predict which paths changed (mass content migration, root-level config swap).
• A security incident may have served compromised content from cache.
• You’re decommissioning a distribution and want a clean slate before final teardown.

​

Cost of a Purge All

​

Console flow

1. Open the distribution → Purge tab.
2. Pick Purge All.
3. Click Purge.

​

Purge via API

curl -X POST "https://api.tenbyte.io/cdn/distributions/$DISTRIBUTION_ID/purge/all" \
  -H "Authorization: Bearer $TENBYTE_API_TOKEN"

{
  "purge_id": "prg_01HXYZ...",
  "status": "queued",
  "scope": "all"
}

curl -sS "https://api.tenbyte.io/cdn/distributions/$DISTRIBUTION_ID/purge/$PURGE_ID" \
  -H "Authorization: Bearer $TENBYTE_API_TOKEN" | jq '.status'

​

Pre-flight checklist

• Origin can absorb a cold-start traffic spike (autoscale ready or capacity proven).
• No live deploy in progress.
• Stakeholders notified (#oncall, #web-platform).
• Confirmed pattern purge cannot solve the problem.
• Have a rollback plan if origin fails to handle the spike.

​

Verify

# Before — typical HIT ratio
curl -sSI "$CDN_HOST/" | grep -i x-cache
# x-cache: HIT

# Trigger purge, then any path
curl -sSI "$CDN_HOST/" | grep -i x-cache
# x-cache: MISS

# Watch [cache hit ratio] dashboard recover over the next few minutes.

​

Recovery

• Watch Cache hit ratio — it nosedives immediately, recovers as paths warm.
• Watch origin error rate. If origin starts to fail, throttle traffic at your edge or scale up.
• Don’t trigger another purge while recovery is in flight.

​

Alternatives