Self hosted smtp monitor

gbzret4d · March 2025

Hi,
what are good self hosted or free smtp monitors? If free I need atleast the possibility to check 20 smtp servers at the same time. The smtp servers are not self hosted so I don't have access to the server or the logs itself.

DrNutella · April 2025

I only have used Nagios Zabbix and Prometheus. Many of my clients use Prometheus.

I hear Incinga is good but didn’t try yet

MannDude · April 2025

Librenms

faleddo · April 2025

Why not selfhost UptimeKuma?

hyperblast · April 2025

for monitoring any service, i recommend @CharityHost_org . he has extensive experience with this.

jschroeder · April 2025

tbh i think it would be easier to write a python script that attempts to log in to smtp servers. Or use telnet for a simpler approach.

nqservices · April 2025

@faleddo said:
Why not selfhost UptimeKuma?

Can uptimekuma monitor smtp with login authentication?

gbzret4d · April 2025

@hyperblast said:
for monitoring any service, i recommend @CharityHost_org . he has extensive experience with this.

So just save some porn mails?

gbzret4d · April 2025

@MannDude said:
Librenms

Thanks, didn't knew librenms is supporting smtp monitoring

MannDude · April 2025

@gbzret4d said:

@MannDude said:
Librenms

Thanks, didn't knew librenms is supporting smtp monitoring

Sorry, was half awake when I posted this. I read this as SNMP monitor.

quicksilver03 · April 2025

If you're OK with self hosting the Prometheus stack (Prometheus, Alertmanager, Blackbox Exporter), the following blackbox-exporter configuration does the job very well:

  smtp:
    prober: "tcp"
    timeout: "30s"
    tcp:
      preferred_ip_protocol: "ip4"
      query_response:
        - expect: "^220 (.*) ESMTP (.+)$"
        - send: "EHLO blackboxexporter.ptrdns.net\r"
        - expect: "^250-STARTTLS"
        - send: "QUIT"

kait · April 2025

@MannDude said: Sorry, was half awake when I posted this. I read this as SNMP monitor.

Sammeee XD

COLBYLICIOUS · April 2025

HetrixTools?

https://docs.hetrixtools.com/v0-9-34-smtp-monitoring/
https://docs.hetrixtools.com/add-a-smtp-monitor/

CloudHopper · April 2025

@quicksilver03 said:
If you're OK with self hosting the Prometheus stack (Prometheus, Alertmanager, Blackbox Exporter), the following blackbox-exporter configuration does the job very well:
  smtp:
    prober: "tcp"
    timeout: "30s"
    tcp:
      preferred_ip_protocol: "ip4"
      query_response:
        - expect: "^220 (.*) ESMTP (.+)$"
        - send: "EHLO blackboxexporter.ptrdns.net\r"
        - expect: "^250-STARTTLS"
        - send: "QUIT"

I'm using Blackbox Exporter for all my uptime monitoring and I really like it. You can run a Prometheus stack with Grafana using docker compose, and if you add Node Exporters to all your servers you'll also get their usage data, (for CPU/RAM consumption etc etc etc).

It's a bit of effort to get it setup but it's well worth it when you have it all up and running.

wadhah · April 2025

@CloudHopper said:
@quicksilver03 said:
If you're OK with self hosting the Prometheus stack (Prometheus, Alertmanager, Blackbox Exporter), the following blackbox-exporter configuration does the job very well:
  smtp:
    prober: "tcp"
    timeout: "30s"
    tcp:
      preferred_ip_protocol: "ip4"
      query_response:
        - expect: "^220 (.*) ESMTP (.+)$"
        - send: "EHLO blackboxexporter.ptrdns.net\r"
        - expect: "^250-STARTTLS"
        - send: "QUIT"
I'm using Blackbox Exporter for all my uptime monitoring and I really like it. You can run a Prometheus stack with Grafana using docker compose, and if you add Node Exporters to all your servers you'll also get their usage data, (for CPU/RAM consumption etc etc etc).

It's a bit of effort to get it setup but it's well worth it when you have it all up and running.

You also need 32Tb ram for this setup

Very lightweight

CloudHopper · April 2025

@wadhah said:
@CloudHopper said:
@quicksilver03 said:
If you're OK with self hosting the Prometheus stack (Prometheus, Alertmanager, Blackbox Exporter), the following blackbox-exporter configuration does the job very well:
  smtp:
    prober: "tcp"
    timeout: "30s"
    tcp:
      preferred_ip_protocol: "ip4"
      query_response:
        - expect: "^220 (.*) ESMTP (.+)$"
        - send: "EHLO blackboxexporter.ptrdns.net\r"
        - expect: "^250-STARTTLS"
        - send: "QUIT"
I'm using Blackbox Exporter for all my uptime monitoring and I really like it. You can run a Prometheus stack with Grafana using docker compose, and if you add Node Exporters to all your servers you'll also get their usage data, (for CPU/RAM consumption etc etc etc).

It's a bit of effort to get it setup but it's well worth it when you have it all up and running.
You also need 32Tb ram for this setup

Very lightweight

I'm monitoring 30+ servers on a 2C/4G VPS using this stack. In the last 24hrs CPU consumption hasn't exceeded 30%, (averaging less than 10%), RAM consumption is stable at 2.7G and Network I/O averages 750Kbps, (maxing out at ~1Mbps)

wadhah · April 2025

@CloudHopper said:
@wadhah said:
@CloudHopper said:
@quicksilver03 said:
If you're OK with self hosting the Prometheus stack (Prometheus, Alertmanager, Blackbox Exporter), the following blackbox-exporter configuration does the job very well:
  smtp:
    prober: "tcp"
    timeout: "30s"
    tcp:
      preferred_ip_protocol: "ip4"
      query_response:
        - expect: "^220 (.*) ESMTP (.+)$"
        - send: "EHLO blackboxexporter.ptrdns.net\r"
        - expect: "^250-STARTTLS"
        - send: "QUIT"
I'm using Blackbox Exporter for all my uptime monitoring and I really like it. You can run a Prometheus stack with Grafana using docker compose, and if you add Node Exporters to all your servers you'll also get their usage data, (for CPU/RAM consumption etc etc etc).

It's a bit of effort to get it setup but it's well worth it when you have it all up and running.
You also need 32Tb ram for this setup

Very lightweight
I'm monitoring 30+ servers on a 2C/4G VPS using this stack. In the last 24hrs CPU consumption hasn't exceeded 30%, (averaging less than 10%), RAM consumption is stable at 2.7G and Network I/O averages 750Kbps, (maxing out at ~1Mbps)

I mean yeah I was just exaggerating, but any Prometheus stack will be very heavy compared to the rest of the options.

Although it is worth it, grafana is pretty

gbzret4d · April 2025

@MannDude said:
Librenms

Thanks, didn't knew librenms is supporting smtp monitoring> @COLBYLICIOUS said:

HetrixTools?

https://docs.hetrixtools.com/v0-9-34-smtp-monitoring/
https://docs.hetrixtools.com/add-a-smtp-monitor/

Not free to monitor 20 and more smtp servers

gbzret4d · April 2025

@wadhah said:
@CloudHopper said:
@wadhah said:
@CloudHopper said:
@quicksilver03 said:
If you're OK with self hosting the Prometheus stack (Prometheus, Alertmanager, Blackbox Exporter), the following blackbox-exporter configuration does the job very well:
  smtp:
    prober: "tcp"
    timeout: "30s"
    tcp:
      preferred_ip_protocol: "ip4"
      query_response:
        - expect: "^220 (.*) ESMTP (.+)$"
        - send: "EHLO blackboxexporter.ptrdns.net\r"
        - expect: "^250-STARTTLS"
        - send: "QUIT"
I'm using Blackbox Exporter for all my uptime monitoring and I really like it. You can run a Prometheus stack with Grafana using docker compose, and if you add Node Exporters to all your servers you'll also get their usage data, (for CPU/RAM consumption etc etc etc).

It's a bit of effort to get it setup but it's well worth it when you have it all up and running.
You also need 32Tb ram for this setup

Very lightweight
I'm monitoring 30+ servers on a 2C/4G VPS using this stack. In the last 24hrs CPU consumption hasn't exceeded 30%, (averaging less than 10%), RAM consumption is stable at 2.7G and Network I/O averages 750Kbps, (maxing out at ~1Mbps)
I mean yeah I was just exaggerating, but any Prometheus stack will be very heavy compared to the rest of the options.

Although it is worth it, grafana is pretty

Too much and too complicated for me. Just want to create my own statistic for years for all email providers I'm using just for myself.

adly · April 2025

@gbzret4d said:
@wadhah said:
@CloudHopper said:
@wadhah said:
@CloudHopper said:
@quicksilver03 said:
If you're OK with self hosting the Prometheus stack (Prometheus, Alertmanager, Blackbox Exporter), the following blackbox-exporter configuration does the job very well:
  smtp:
    prober: "tcp"
    timeout: "30s"
    tcp:
      preferred_ip_protocol: "ip4"
      query_response:
        - expect: "^220 (.*) ESMTP (.+)$"
        - send: "EHLO blackboxexporter.ptrdns.net\r"
        - expect: "^250-STARTTLS"
        - send: "QUIT"
I'm using Blackbox Exporter for all my uptime monitoring and I really like it. You can run a Prometheus stack with Grafana using docker compose, and if you add Node Exporters to all your servers you'll also get their usage data, (for CPU/RAM consumption etc etc etc).

It's a bit of effort to get it setup but it's well worth it when you have it all up and running.
You also need 32Tb ram for this setup

Very lightweight
I'm monitoring 30+ servers on a 2C/4G VPS using this stack. In the last 24hrs CPU consumption hasn't exceeded 30%, (averaging less than 10%), RAM consumption is stable at 2.7G and Network I/O averages 750Kbps, (maxing out at ~1Mbps)
I mean yeah I was just exaggerating, but any Prometheus stack will be very heavy compared to the rest of the options.

Although it is worth it, grafana is pretty
Too much and too complicated for me. Just want to create my own statistic for years for all email providers I'm using just for myself.

In that case, are you looking for actual SMTP monitoring or would monitoring SMTP port(s) be sufficient? There's likely to be many more options if you just want to check the service/port is online.

wadhah · April 2025

@gbzret4d said:
@wadhah said:
@CloudHopper said:
@wadhah said:
@CloudHopper said:
@quicksilver03 said:
If you're OK with self hosting the Prometheus stack (Prometheus, Alertmanager, Blackbox Exporter), the following blackbox-exporter configuration does the job very well:
  smtp:
    prober: "tcp"
    timeout: "30s"
    tcp:
      preferred_ip_protocol: "ip4"
      query_response:
        - expect: "^220 (.*) ESMTP (.+)$"
        - send: "EHLO blackboxexporter.ptrdns.net\r"
        - expect: "^250-STARTTLS"
        - send: "QUIT"
I'm using Blackbox Exporter for all my uptime monitoring and I really like it. You can run a Prometheus stack with Grafana using docker compose, and if you add Node Exporters to all your servers you'll also get their usage data, (for CPU/RAM consumption etc etc etc).

It's a bit of effort to get it setup but it's well worth it when you have it all up and running.
You also need 32Tb ram for this setup

Very lightweight
I'm monitoring 30+ servers on a 2C/4G VPS using this stack. In the last 24hrs CPU consumption hasn't exceeded 30%, (averaging less than 10%), RAM consumption is stable at 2.7G and Network I/O averages 750Kbps, (maxing out at ~1Mbps)
I mean yeah I was just exaggerating, but any Prometheus stack will be very heavy compared to the rest of the options.

Although it is worth it, grafana is pretty
Too much and too complicated for me. Just want to create my own statistic for years for all email providers I'm using just for myself.

Honestly it might be your best option if you want the most customizable option. I understand it can be rough at the start (their docs SUCK) but you just set it and forget it after the first setup, it's prod ready and very resilient .

gbzret4d · April 2025

@wadhah said:
@gbzret4d said:
@wadhah said:
@CloudHopper said:
@wadhah said:
@CloudHopper said:
@quicksilver03 said:
If you're OK with self hosting the Prometheus stack (Prometheus, Alertmanager, Blackbox Exporter), the following blackbox-exporter configuration does the job very well:
  smtp:
    prober: "tcp"
    timeout: "30s"
    tcp:
      preferred_ip_protocol: "ip4"
      query_response:
        - expect: "^220 (.*) ESMTP (.+)$"
        - send: "EHLO blackboxexporter.ptrdns.net\r"
        - expect: "^250-STARTTLS"
        - send: "QUIT"
I'm using Blackbox Exporter for all my uptime monitoring and I really like it. You can run a Prometheus stack with Grafana using docker compose, and if you add Node Exporters to all your servers you'll also get their usage data, (for CPU/RAM consumption etc etc etc).

It's a bit of effort to get it setup but it's well worth it when you have it all up and running.
You also need 32Tb ram for this setup

Very lightweight
I'm monitoring 30+ servers on a 2C/4G VPS using this stack. In the last 24hrs CPU consumption hasn't exceeded 30%, (averaging less than 10%), RAM consumption is stable at 2.7G and Network I/O averages 750Kbps, (maxing out at ~1Mbps)
I mean yeah I was just exaggerating, but any Prometheus stack will be very heavy compared to the rest of the options.

Although it is worth it, grafana is pretty
Too much and too complicated for me. Just want to create my own statistic for years for all email providers I'm using just for myself.
Honestly it might be your best option if you want the most customizable option. I understand it can be rough at the start (their docs SUCK) but you just set it and forget it after the first setup, it's prod ready and very resilient .

I will set up a test server.... I've got enough idling server. I'm just unsure how much space it will need if I'm really trying to run it for years...

gbzret4d · April 2025

@adly said:
@gbzret4d said:
@wadhah said:
@CloudHopper said:
@wadhah said:
@CloudHopper said:
@quicksilver03 said:
If you're OK with self hosting the Prometheus stack (Prometheus, Alertmanager, Blackbox Exporter), the following blackbox-exporter configuration does the job very well:
  smtp:
    prober: "tcp"
    timeout: "30s"
    tcp:
      preferred_ip_protocol: "ip4"
      query_response:
        - expect: "^220 (.*) ESMTP (.+)$"
        - send: "EHLO blackboxexporter.ptrdns.net\r"
        - expect: "^250-STARTTLS"
        - send: "QUIT"
I'm using Blackbox Exporter for all my uptime monitoring and I really like it. You can run a Prometheus stack with Grafana using docker compose, and if you add Node Exporters to all your servers you'll also get their usage data, (for CPU/RAM consumption etc etc etc).

It's a bit of effort to get it setup but it's well worth it when you have it all up and running.
You also need 32Tb ram for this setup

Very lightweight
I'm monitoring 30+ servers on a 2C/4G VPS using this stack. In the last 24hrs CPU consumption hasn't exceeded 30%, (averaging less than 10%), RAM consumption is stable at 2.7G and Network I/O averages 750Kbps, (maxing out at ~1Mbps)
I mean yeah I was just exaggerating, but any Prometheus stack will be very heavy compared to the rest of the options.

Although it is worth it, grafana is pretty
Too much and too complicated for me. Just want to create my own statistic for years for all email providers I'm using just for myself.
In that case, are you looking for actual SMTP monitoring or would monitoring SMTP port(s) be sufficient? There's likely to be many more options if you just want to check the service/port is online.

SMTP ports would be enough

rcy026 · April 2025

@gbzret4d said:

SMTP ports would be enough

Then Uptime Kuma will probably be enough.

gbzret4d · April 2025

@rcy026 said:

@gbzret4d said:

SMTP ports would be enough

Then Uptime Kuma will probably be enough.

What would you suggest if checking the port would not be sufficient?

rcy026 · April 2025

@gbzret4d said:

@rcy026 said:

@gbzret4d said:

SMTP ports would be enough

Then Uptime Kuma will probably be enough.

What would you suggest if checking the port would not be sufficient?

If you wanna go all in, for monitoring I would go with Nagios, Icinga or CheckMK, store the data in a Prometheus or Influx and visualize it with Grafana.
I would monitor the port itself with a simple tcp check and then make a service that sends an email using the specified smtp and checks that the email is delivered all the way via pop or imap.
If it was my smtp server I would keep track of the vitals with node exporter.

oplink · April 2025

Check out librenms, it can work with Grafana/Prometheus if you want too

librenms is very easy to get setup too

gbzret4d · April 2025

@oplink said:
Check out librenms, it can work with Grafana/Prometheus if you want too

librenms is very easy to get setup too

I already tried librenms and the learning curve is not that hard. I didn't knew that smtp monitoring is possible with librenms.

aglodek · April 2025

Ansible?

There's a learning curve, like with most things, but once you take control, you can use it for pretty much anything!

greenhost_cloud · April 2025

You might want to check out tools like SMTPDiag or MailHogm.

gbzret4d · April 2025

@greenhost_cloud said:
You might want to check out tools like SMTPDiag or MailHogm.

Thanks, will take a look at both. Never heard about them.

Howdy, Stranger!

Categories

In this Discussion

Self hosted smtp monitor

Comments

Howdy, Stranger!

Quick Links

Categories

In this Discussion

Self hosted smtp monitor

Comments