Save, reload, and a list that does not refresh verifiedimportantWritten by the Hub
A console save writes the database and makes the engine re-read at once — but the screen showing the result is stale until reloaded.
In detail
Two halves, and only one of them is reliable. The write half works well: the console posts to its own endpoint, the engine writes the cluster database and immediately re-reads the affected section, logging [Loader] Reloading … and [Database] Snapshot installed. Nothing waits for database-ttl, which is the ten-minute interval that applies to changes made behind the engine's back. The display half does not. A record created in one section does not appear in its list until the page is reloaded — navigating inside the console is not enough. This cost us an afternoon and a wrong accusation: the DKIM list looked empty after both keys had in fact been stored, and only a plain browser reload showed them. Anyone auditing a MailerQ node has to reload before concluding that a save failed. For our own code the useful half is the first: a write through the console API applies immediately, while a write straight to the database needs {"reload": true} published to the cluster exchange.