Haraka VS sanoid

postale.io has an MX record that points to mail.postale.io, which resolves to an AWS IP address (probably an EC2 instance) and has Haraka listening on port 25 (if its SMTP banner can be believed)

I found that Zerodha uses Haraka and Karan–their DevOps guy, mentioned it is easy-setup and performant.

The next time this happens i'll move to a self hosted solution, like haraka or anonaddy

Haraka Like: Simple to hook into the processing pipeline Dislike: Node.js and had some issues with stability when I last tried it

Having another SMTP as a choice is cool but not as fun as something that is way more configurable (ex. Haraka), or API driven (a la Postal).

Look at Haraka: http://haraka.github.io/

For the record, I'm using dovecot (https://www.dovecot.org/) for IMAP and Haraka (https://haraka.github.io/) for SMTP.

I run my own mail servers for small projects, though for my main email I've actually switched to ProtonMail.

It's never been easier to self host your email with projects like the following around:

- https://foxcpp.dev/maddy/

- https://github.com/albertito/chasquid

- https://github.com/haraka/haraka

Of course the usual dovecot + postfix setup is great for learning even if a bit complicated.

Haraka (https://haraka.github.io/) is a plugin-based mailserver written in JavaScript. I've messed with postfix before but found Haraka easier because instead of arcane configuration files I can just read, copy, and tweak a plugin to suit my needs. Policies for receiving, storing, forwarding, and ignoring mail can be arbitrarily complex.

I’m on the other end of the spectrum. I like knowing the flags and settings I use to create the pools.

For snapshots and replication take a look at sanoid (https://github.com/jimsalterjrs/sanoid).

I use a combo of PBS and ZFS with Sanoid. The ZFS snapshots are nice because you define it all in a config file with frequent, hourly, daily, weekly, monthly snaps retained however you want. Importantly I use that on the root pool as well as my data pools. With PBS I have a custom retention config and have it backing up VMs and LXCs every couple hours. The PBS dedupe functionality makes it possible to keep full backups going back ages with minimal storage consumption, just the changed blocks. PBS is running inside a VM yes, but it’s storage is on another pool.

[3] https://github.com/jimsalterjrs/sanoid/

No, it's written by the developer of Sanoid and Syncoid. The foremost open source ZFS snapshotting tool. He contributes on this sub all the time.

Hi there! Not so much to answer your questions, but are you aware of Sanoid and syncoid? It does backup and remote Sync in the most seamless way. Perhaps this would solve some of your concerns...