IIUC event MPM can close keepalive connections up to 100ms than the configured value.
If other software parses the Keep-Alive response header and fudges the result by e.g. tens of milliseconds for its own TTL to avoid races, should we internally try to avoid these whole-second keepalivetimeouts? For example under event adding the 100ms back in so at worst the timeout occurs in the same timeout=X advertised? In the case I was working on, the client was doing no fudging on the whole seconds returned by timeout=X so it's connection cache TTL can easily get things wrong. A promising workaround is to configure e.g. 5999ms which will advertise timeout=5 but wait close to 6 seconds. -- Eric Covener cove...@gmail.com
