fix(customer-routes): Cache-Control: no-store on customer endpoints (#470)

The trigger: PR #458 mounted requireCustomerPortalEnabled which
410'd every /api/customer/* + /api/admin/customers/* request when
the master toggle was off. Some browsers cached that 410 (no
Cache-Control header was set, so heuristic freshness applied —
the wrong default for an authenticated/sensitive surface).
PR #470 reverted the middleware, but a customer whose tab cached
the 410 still saw 410s until they hard-refreshed.

Add noStoreCache middleware and mount it in front of both route
groups. Every response (200, 4xx, 5xx) now carries
`Cache-Control: no-store, no-cache, must-revalidate, private`
plus the HTTP/1.0 Pragma + Expires fallbacks. Any future
transient error from these endpoints can no longer get pinned in
browser or proxy caches and outlive its cause.

Cost is one setHeader per request; applied per route group rather
than globally so static assets + galleries keep their own caching
strategy unchanged.

Includes a dedicated unit test pinning the header set so a future
cleanup pass can't quietly drop it and re-introduce the bug.
This commit is contained in:
Paul Nothaft
2026-05-12 22:53:49 +02:00
parent 5e86eef4f8
commit 3122dd08a8
3 changed files with 98 additions and 3 deletions
+39
View File
@@ -0,0 +1,39 @@
/**
* Cache-Control: no-store helper for sensitive endpoints.
*
* Why a dedicated middleware: shipping the wrong cache-control header
* on a session-bearing endpoint is a class-of-bug that bites long
* after the original mistake. The PR #458 / PR #470 history is the
* concrete trigger:
*
* - #458 mounted requireCustomerPortalEnabled which 410'd every
* /api/customer/* and /api/admin/customers/* request when the
* master toggle was off.
* - Some browsers cached the 410 (the response carried no explicit
* Cache-Control header, so heuristic freshness applied — for an
* authenticated/sensitive surface that's the wrong default).
* - #470 reverted the middleware, but a customer whose tab cached
* the 410 still saw 410s until they hard-refreshed.
*
* Mounting `noStoreCache` in front of these routes belt-and-braces
* the future: any 4xx/5xx (or 200) response from these endpoints
* carries `Cache-Control: no-store`, so a transient kill-switch,
* permission flip, or backend restart can never get pinned in
* intermediate caches.
*
* No-op cost (one setHeader per request); applied per route group
* rather than globally so static assets + galleries keep their
* own caching strategy.
*/
function noStoreCache(req, res, next) {
// `no-store` is the strongest signal — no cache, no revalidation,
// no offline retention. Pair with `private` so any well-behaved
// intermediate proxy treats the response as user-specific.
res.setHeader('Cache-Control', 'no-store, no-cache, must-revalidate, private');
res.setHeader('Pragma', 'no-cache'); // HTTP/1.0 fallback for older proxies
res.setHeader('Expires', '0');
next();
}
module.exports = { noStoreCache };