<domain>set by the Hub
The same as max-smtp-out, but it applies while the queue is in backoff. Connections are the first thing worth cutting when a receiver starts refusing: it watches them more closely than volume.
Usually one or two is set. The point is not to deliver more slowly but to stop looking pushy while the issue is being sorted out.
| Mode | Per address (vMTA) | In the provider rule | When this is your case |
|---|---|---|---|
| Warm-up, cold address | 1 | — | An address younger than 4–6 weeks or back from a pause |
| Careful | 1 | — | Strict receivers, questionable reputation, the first weeks after warm-up |
| Balanced default | 2 | — | Normal operation of an established fleet |
| Aggressive | 5 | — | Warmed-up addresses, friendly receivers, urgent volumes |
| PowerMTA default | 20 | — | If the directive is not written at all |
Single digits, not dozens. A receiving server counts connections most carefully of all: a queue that slowed down because of it should not keep as many conversations with it as before.
It has its own default — 20, unlike the other directives in the family, which take the normal value. Twenty connections for backoff mode is a lot, and leaving the default is almost always wrong.
Count it together with the rate. One message an hour does not need even one permanent connection: two or three are more than enough.
# in backoff mode we keep single-digit connections, not dozens <domain hotmail.com> max-smtp-out 15 backoff-max-smtp-out 2</domain>
what people get wrong and how it ends2 mistakes▶
§ 3.3.10.85
our experience
2 relations
pmta show settings hotmail.com/vmta1 | grep -i backoff-max-smtp-out pmta show queues --mode=backoff --maxitems=20
3 articles
◇Four value types the whole config is written inBoolean, domain pattern, regular expression and time interval — defined once in § 3.2 and then used by all directives.◇How the policy for a receiver is assembledSettings are assembled from the most general pattern to the most precise; the order of blocks in the file means nothing.◇Backoff mode — the queue's second personalityA queue that a receiver has started refusing switches to a parallel set of limits, retries and routes — ten directives that act only in this mode.