keepalive: a ping that missed and re-wrote the cache does not restart the ttl_expiry clock #580

Summary

A keep-alive ping that missed the cache does not move the entry's refresh time. The next miss can then get the verdict ttl_expiry (the cache lifetime ran out) when the cache was in fact re-created by that ping. Found by Osher in the review of #546 (finding 4).

What we found

• proxy/keepalive.go:1538 and :1776 move the refresh time only for a ping that read the cache (CacheRead > 0).
• A ping that missed also wrote the entry again, so the cache lifetime started again at that ping.
• The next request that misses is then judged against the older refresh time, and the dashboard shows ttl_expiry.
• This is rare at the default 270 s idle setting.

Proposal

1. Move the refresh time for a ping that wrote the cache (CacheWrite > 0) as well as one that read it.
2. Add a test: a ping that misses and writes, then a request after the lifetime from the first request but within the lifetime from the ping. The verdict must not be ttl_expiry.

Reference: #546 (comment)