Pantheon next-gen Global CDN: self-upgrade by March 31, 2027—or wait for the forced move
Pantheon's October 2026 release note sets March 31, 2027 as the self-serve deadline to move to next-gen Global CDN on Cloudflare. From April 2027 Pantheon migrates what remains, with 60+ days' notice—and AGCDN customers must wait for Pantheon.

Pantheon's release note (October 26, 2026, edited October 29) asks sites still on classic Global CDN to upgrade to the next-gen Global CDN (Cloudflare) by March 31, 2027. From April 2027 Pantheon migrates remaining sites with at least sixty days' notice. Bot protection turns on by default on next-gen. Advanced Global CDN (AGCDN) customers should not self-migrate—contact Pantheon by December 15, 2026 and wait for their path. Sites created from September 10, 2026 are already on next-gen. Plan cache, WAF and bot-exception tests before you click the dashboard banner.
What the release note actually says
Pantheon published a release note dated October 26, 2026 (edited October 29) about upgrading to the next-generation Global CDN built on Cloudflare. Sites still on the classic Global CDN are asked to self-upgrade by March 31, 2027. Starting April 2027, Pantheon will migrate remaining sites and will give at least sixty days' notice. The Site Dashboard shows a GCDN banner for eligible properties. Bot protection is on by default on next-gen. Sites created on or after September 10, 2026 are already on next-gen.
Advanced Global CDN (AGCDN) is a different path: Pantheon tells those customers not to self-migrate and to contact Pantheon by December 15, 2026 so the platform team can handle the move. If you are on AGCDN and you click around looking for the same self-serve button as GCDN, stop—follow the AGCDN instruction instead.
The dashboard banner is a calendar invite with blast radius. Treat it like a cutover, not a checkbox.
Who should act on which date
Already on next-gen (sites from September 10, 2026 onward). Confirm in the dashboard. Spot-check cache headers and bot behavior anyway—defaults may differ from what your older runbooks assume.
Classic GCDN, self-serve eligible. Pick a maintenance window before March 31, 2027. Do not wait for the April forced wave unless you like scheduling around Pantheon's notice window.
AGCDN. Contact Pantheon by December 15, 2026. Build your test plan now so you are ready when they schedule you. Your custom edge rules are why you bought AGCDN; they are also why a casual click is wrong.
Multi-site portfolios. Inventory every Pantheon site: GCDN vs AGCDN, last cache purge procedure, owners, and whether marketing has a big launch pinned to Q1 2027. Stagger upgrades so one bad bot rule does not take every property in the same afternoon.
Decision guide
Who acts on which date
| Consideration | Site type | Action |
|---|---|---|
| Next-gen already | Sites from Sep 10, 2026. | Spot-check defaults |
| Classic GCDN | Self-upgrade by Mar 31, 2027. | Or wait for Apr+ wave |
| AGCDN | Do not self-migrate. | Contact by Dec 15, 2026 |
| Forced wave | From April 2027. | 60+ days' notice |
The dashboard banner is a calendar invite with blast radius. Treat it like a cutover, not a checkbox.
The CDN banner as a cutover
Before anyone clicks upgrade on a production site, run this checklist.
- Snapshot the current edge behavior. Cache TTLs, cookie-bypass rules, redirect chains, image optimizer settings, WAF exceptions, IP allows for monitoring and agency tools.
- List bot-sensitive flows. Checkout, login, form POSTs, preview URLs, headless CMS previews, synthetic monitors. Bot protection on by default will false-positive something you forgot.
- Stage first. Upgrade a non-prod or low-traffic site. Compare HTML, asset headers and form posts to a baseline capture.
- Purge and warm. Know who may purge, how long warm takes for your top paths and what "site looks empty" means in the first ten minutes.
- Rollback criteria. Write the symptoms that mean "open a Pantheon ticket / roll back" before the window starts—not in the Slack thread during the window.
- CMS and plugin owners. Drupal and WordPress cache modules, reverse-proxy assumptions and CDN modules sometimes hard-code old header expectations. Have those owners on the bridge.
This is ordinary web application cutover hygiene. The Cloudflare under the hood does not remove the need for it.
Where portfolios get surprised
AGCDN teams self-serve by mistake. December 15 exists so Pantheon can plan custom configurations. Use it.
Bot protection vs. partners. Agencies, uptime checkers and translation crawlers often need allowlists. Collect those IPs before the default turns on.
Cached error pages. A next-gen platform that caches 500s more aggressively than you expect will show a bad page to more of the world. Confirm error caching policy in staging.
Launch calendars. A March 31 self-serve deadline that slips into April collides with spring campaigns. Book the window in January if your Q1 is crowded.
"We are on Pantheon so CDN is fine." Platform defaults are not your runbook. The release note is the prompt to refresh the runbook.
Related reading when edge behavior meets product delivery: step functions as agent tools is a different layer, but the same lesson applies—platform features still need an owner and a test matrix.
A ninety-day portfolio plan
Now through December 15, 2026: AGCDN customers contact Pantheon; GCDN owners finish the inventory and pick pilot sites.
January 2027: Upgrade pilots; fix bot false positives; document purge/warm.
February 2027: Roll the rest of self-serve GCDN sites; keep a freeze list for anything launching that month.
By March 31, 2027: Self-serve complete or explicitly waiting for Pantheon's April+ wave with the sixty-day notice on the calendar.
April 2027 onward: Track Pantheon notices; do not discover a forced migrate in a marketing freeze.
Sequence of work
Ninety-day portfolio plan
- By Dec 15
AGCDN contacts Pantheon.
- January
Pilot GCDN upgrades.
- February
Roll remaining self-serve.
- By Mar 31
Self-serve done or freeze listed.
If the banner is on your dashboard
Inventory GCDN vs AGCDN this week. AGCDN: contact Pantheon before December 15. Everyone else: schedule a real cutover before March 31, 2027, with bot and cache tests first. If you want help turning the banner into a portfolio runbook inside web applications, get in touch.
One more practical note before you close the tab. Put the deadline on a shared calendar with a named owner, not a channel that everyone watches and no one owns. Committees do not ship carton changes, agent pilots or CDN cutovers. A single person who can say "green" or "blocked on X" keeps the work from restarting every Monday with the same open questions.
When you brief leadership, lead with the operating decision, not the vendor announcement. Executives do not need another slide about artificial intelligence or carrier math. They need to know what changes at the bench, in checkout, in the partner portal or in the release train—and who is accountable for the before-and-after measurement.
Keep a short evidence packet next to every change: the primary source URL, the date you verified it, the SKUs or environments in scope and the rollback path. That packet is what you bring to the QBR, the board update or the incident review. Memory and Slack threads are not evidence.
If outside help would finish the inventory, the redesign or the cutover faster than waiting for a quiet week that never comes, bring them into the first scoping conversation with the right TWIMCO team rather than after the first red invoice or console banner. Scheduled migrations beat emergency freezes.
Write the "what good looks like" line in the same doc as the deadline. Green means a measured outcome—cartons repriced, keys rotated, banners cleared, agents stuck behind a real gate—not a meeting that ended on time. Share that definition with whoever will inherit the work when you are out.
Resist the urge to expand scope mid-pilot. One SKU family, one ad account, one Pantheon site or one MCP client is enough to learn whether the controls hold. Expansion is a reward for a boring audit log, not a way to feel productive in week one.
Related work: Sutton Place.
We build alongside the people doing the work.
Bring us the problem you’re thinking about.