421 4.7.0 medium
421 4.7.0 Temporary deferral / too many connections / rate limited
A soft bounce: the receiving server is throttling you for sending too fast or opening too many connections, and asks you to slow down and retry.
What you see
421 4.7.0 Try again later, closing connection 421-4.7.0 [203.0.113.5] Our system has detected an unusual rate 421 4.7.0 of unsolicited mail originating from your IP address.
What’s actually happening
Mid-send the receiver returns a 421 and drops the connection. The 4.x.x class means temporary — the message isn't rejected, just deferred, and a well-behaved sender retries later and gets through. It shows up when you blast a large batch quickly, open dozens of parallel connections, or your reputation dipped enough that the provider wants you to throttle. Ignore the signal and keep hammering, and the soft 421 can harden into a 5.7.1 block.
Common causes
- Sending rate exceeds the provider's per-IP/per-domain limit (Gmail, Outlook, Yahoo each cap messages per connection and per hour)
- Too many concurrent SMTP connections from one IP — providers cap simultaneous sessions
- Greylisting: the server defers an unknown sender on first contact and expects a retry
- Reputation/content flag — borderline spam signals trigger throttling rather than an outright block
- A burst from a new or cold IP with no established sending history
How to fix it
- Back off and retry with exponential delay421 means try again, not give up. Retry the deferred message after a growing delay (e.g. 5 min, then 15, 30, 60) rather than immediately. A proper mail queue (Postfix, or your ESP) does this automatically; a naive script that retries instantly just gets deferred again and looks abusive.
- Cut concurrency and send rateReduce simultaneous connections to a handful and add a small pause between messages so you stay under the provider's per-hour ceiling. For bulk, throttle to the documented limits (Gmail and Outlook publish theirs) instead of firing everything at once.
- Honor greylistingIf the defer happens on first contact and clears on retry minutes later, that's greylisting working as designed — your retry queue handles it with no change needed. Confirm your sender actually retries; one-shot scripts silently drop greylisted mail.
- Check reputation if 421s persistRepeated throttling from many providers means a reputation problem, not just speed. Check Google Postmaster Tools and blocklists, authenticate with SPF/DKIM/DMARC, and warm a cold IP gradually. Persistent 421s are the warning shot before a hard block.
- Route bulk through a dedicated relaySending large volume directly from a web/shared host invites throttling. Use SES/Postmark/SendGrid, which manage IP reputation, connection pooling, and rate limits for you.
Stop it recurring
Send through a queue that retries with backoff and respects per-provider rate limits instead of blasting batches from a single connection.
Related errors