SMTP 421 Service Not Available: Causes and Fixes

A 421 reply means the server cannot take your message right now and is hanging up. The leading 4 marks a transient failure, so your mail server should queue the message and retry on its own. The text after the number decides whether this is routine throttling or the first sign of a reputation problem.

RFC 5321 defines the base text as "Service not available, closing transmission channel". In practice, receivers attach their own explanation, and the enhanced code (4.3.2, 4.4.5, 4.7.0 and so on) tells you which kind of refusal it is. Single-code lookups for the 4xx family are in email bounce codes explained; this page is the deep dive for 421 only.

postfix/smtp[18211]: 3F2A9C1B04: to=<alice@example.net>, relay=mx.example.net[198.51.100.20]:25, delay=1.4, dsn=4.7.0, status=deferred (host mx.example.net[198.51.100.20] said: 421 4.7.0 Too many connections from your IP, try again later (in reply to RCPT TO command))

Note the last part: "in reply to RCPT TO command" (or end of DATA, or MAIL FROM) tells you how far the session got before the receiver hung up.

Common 421 variants

Check Gmail's SMTP error reference for exact current wording.

Causes, most likely first

1. Rate limiting by the receiver. You opened too many connections or sent too many messages in a short window.
2. Poor or unknown reputation. New IP, new domain, or a sudden jump in volume.
3. Authentication or DNS gaps. Missing PTR, no SPF/DKIM, no DMARC.
4. The receiver is down or in trouble. Maintenance, overload, a failed disk.
5. Your own server is the one saying 421. Postfix, Exim or your relay can return 421 to your clients when it is out of memory, disk or connection slots.

Step 1: decide whose 421 it is

Look at the relay= host in your log. If it is the recipient's MX, the remote server said it. If the line reads localhost or your own relay, you are debugging your server. Find the recipient's MX with the MX Lookup and confirm the port is reachable with the Port Checker.

sudo grep -E "dsn=4\.[0-9.]+, status=deferred" /var/log/mail.log | grep " 421" | tail -n 20
postqueue -p | tail -n 3       # queue size

Step 2: read the enhanced code and the explanation

If the text names your IP, PTR, authentication, TLS or reputation, fix that item first. A PTR complaint is solved in DNS, see reverse DNS (PTR) for email. An authentication complaint goes to the Email DNS Checker and the SPF Record Checker. Check the IP on the Blacklist Checker too, because a listing often shows up as 421 throttling before it becomes a 554.

Step 3: slow down

If the text is about rate or connections, cap your concurrency toward that destination. In Postfix, add a per-domain transport rule so one slow receiver does not get hammered:

  # /etc/postfix/main.cf
slow_destination_concurrency_limit = 2
slow_destination_rate_delay = 1s
transport_maps = hash:/etc/postfix/transport

  # /etc/postfix/transport
example.net   slow:

sudo postmap /etc/postfix/transport
  # add to master.cf:  slow  unix  -  -  n  -  -  smtp
sudo postfix reload

Postfix's default_destination_rate_delay and smtp_destination_concurrency_limit serve the same purpose globally. Check your version's documentation for the exact parameter names before applying.

Step 4: if your own server returns 421

Check free memory, disk and connection limits. Postfix returns 421 when it cannot accept more clients than smtpd_client_connection_count_limit allows (default 50 per client), and a full /var partition causes queue errors. Watch df -h and free -m, then read /var/log/mail.log around the failures.

How to verify

Do not shorten retry intervals to seconds; the default queue retries for up to maximal_queue_lifetime (5 days). Run postqueue -f to flush once after the fix, then watch for status=sent to that domain. Send a test through the Email Deliverability Test to confirm authentication and PTR are clean.

Next steps

• Review email bounce codes explained for the whole 4xx and 5xx map.
• Check your sending IP on the Blacklist Checker.
• Warm a new sender properly with email warmup for a new domain.
• Read SMTP 450 mailbox unavailable and greylisting for the other common deferral.

FAQ

Is SMTP 421 permanent?

No. It is a transient failure and a normal MTA retries it automatically. Only worry if the same message stays deferred for many hours.

Why did I get 421 from Gmail when my SPF passes?

Gmail may rate limit for volume, reputation, a missing PTR or a failed DKIM or DMARC alignment, not only SPF. Read the exact 4.7.x text in the log.

Should I lower my sending rate?

If the text mentions rate, connections or "unusual rate", yes. Reduce concurrency to that domain and ramp up gradually.

Can 421 come from my own mail server?

Yes. A server out of memory, disk or connection slots can send 421 to its own clients, and so can a relay under maintenance.

This page is general information written by our editorial team and updated when documentation changes. Servers and providers differ, so test changes safely first. See our editorial policy and disclaimer.