---
title: "Hostname"
description: "A hostname is the human-readable name assigned to a device or server on a network, used together with DNS to locate mail servers and other services."
url: "https://mailmum.io/glossary/hostname/"
---

# Hostname

A hostname is the human-readable label assigned to a device or server on a
network, such as `mail.example.com`. Hostnames are resolved to IP addresses
through the Domain Name System (DNS), and back again through
[reverse DNS (rDNS)](https://mailmum.io/glossary/reverse-dns-rdns/) and
[PTR records](https://mailmum.io/glossary/pointer-record-ptr/).

For a mail server the hostname is not cosmetic. It is the identity the server
presents to every other server it talks to, and it is checked.

## What makes a hostname fully qualified?

A fully qualified domain name (FQDN) is a hostname that includes every label up
to the [top-level domain](https://mailmum.io/glossary/top-level-domain-tld/), so it is unambiguous
no matter where it is resolved from:

```txt
mail . example . com
 |       |        |
 |       |        +-- top-level domain
 |       +----------- domain
 +------------------- host label
```

A mail server must use an FQDN &mdash; a bare `mail`, or a name that only
resolves inside your own network, will fail on the receiving side. It also must
resolve publicly: a hostname that only exists in `/etc/hosts` is not a hostname
as far as the internet is concerned.

## Where does the hostname appear in SMTP?

In the greeting. The connecting server announces itself with `HELO` or `EHLO`
followed by its own hostname, before anything else happens:

```txt
$ telnet mx1.example.net 25
220 mx1.example.net ESMTP Postfix
EHLO mail.example.com
250-mx1.example.net
250-PIPELINING
250-SIZE 20480000
250-STARTTLS
250 8BITMIME
```

That name is recorded in the `Received:` header of every message the connection
delivers, which is why it stays visible to the recipient and to any filter that
inspects the header chain:

```txt
Received: from mail.example.com (mail.example.com [192.0.2.25])
	by mx1.example.net (Postfix) with ESMTPS id 4Rk8Zt1234
	for <recipient@example.net>; Tue, 04 Aug 2026 08:11:02 +0000
```

## What is forward-confirmed reverse DNS?

Receiving servers do not take the greeting at face value. They perform a round
trip, and the term for it is forward-confirmed reverse DNS (FCrDNS):

1. Look up the PTR record for the connecting IP address.
2. Look up the A or AAAA record for the hostname that PTR returned.
3. Check that it resolves back to the same IP address.

```sh
$ dig -x 192.0.2.25 +short
mail.example.com.

$ dig a mail.example.com +short
192.0.2.25
```

When those agree, the hostname is confirmed. When they do not &mdash; a missing
PTR record, a PTR pointing at a name with no forward record, or a forward record
resolving somewhere else &mdash; the connection is treated as unverified, and
many receivers reject or heavily penalise it. This is one of the most common
causes of mail from an otherwise correctly configured server being refused.

Ideally all three agree: the `HELO`/`EHLO` name, the PTR record, and the forward
record. They do not strictly have to be identical, but every difference is one
more thing a receiving filter has to be lenient about.

## What makes a bad hostname for a mail server?

* **Generic provider-assigned names.** Anything of the form
  `ip-192-0-2-25.provider.example` or `static.25.2.0.192.clients.example.net`
  reads as unconfigured address space. Some receivers score it, some refuse the
  connection outright.
* **Names suggesting dynamic or residential allocation** &mdash; containing
  `dyn`, `dsl`, `pool`, `dhcp`, `cable`. Several blocklists are built precisely
  around these patterns.
* **A hostname that does not resolve at all,** or one whose forward record
  points somewhere else.
* **An IP address literal in the greeting,** such as `EHLO [192.0.2.25]`. Valid
  syntax, but treated with suspicion.
* **A name in a domain you do not control,** which cannot be aligned with SPF or
  DKIM and looks like an impersonation attempt.

## Why do hostnames matter for blacklisting?

Blocklists and reputation systems key primarily on IP addresses rather than
hostnames, for a simple reason: a hostname can be changed in minutes, while the
IP address is what actually connects and is far harder to replace.

The hostname still matters, because it feeds the checks that happen before any
reputation lookup. A missing or invalid PTR record is frequently enough on its
own to have a connection refused, and a generic hostname raises the score of
every message from that server. See
[IP blacklisting](https://mailmum.io/glossary/ip-blacklisting/) for how the listings themselves
work, and [PTR record](https://mailmum.io/glossary/pointer-record-ptr/) for who controls the
reverse entry and how to get it set.

## Where does this come up in practice?

* [Howto block mailservers using uncommon TLDs in hostname through
  Postfix](https://mailmum.io/posts/howto-block-mailservers-using-uncommon-tlds-in-hostname-through-postfix/)
  uses the greeting hostname as a filtering criterion, and shows why the check
  belongs after `permit_sasl_authenticated`.
* [Discover mail system setup errors using MailMum
  monitoring](https://mailmum.io/posts/discover-mail-system-setup-errors-using-mailmum-monitoring/)
  walks through a real Postfix rejection caused by a HELO name that does not
  match the connecting server's reverse DNS.
