Ability to set IMAPS connector pull faster than 5 min

Support
1

NethServer Version: latest
Module: mail

I have some time critical email, that the minimum set by NS for the IMAPS connector, 5 minutes, is not going to cut it.

Can I configure this (editing some config file or cron job or something) to be faster? (like every minute)

2

Maybe the ISP/system is holding your messages won’t honor your reduced timing of imap checks. Like using 120mm water pipes into your house, when water delivery service use only 14mm pipe. Won’t change a thing.

Email was never meant to be a real time (or close to) message delivery system, so maybe the email should not be the way for deliver comunications?
Anyway…

If you wish to lower the timeout of the communication, maybe you can trigger a conditioned forward to another mailbox only for this purpose, maybe one with push support.

1 Like
3

You can define the minutes in config db:

2 Likes
4

With one little amendment:
It will spawn only one instance of get-mail, (/usr/bin/flock in the cron);
If the last transaction has not finished, it will not start a new one. (no mater how short the interval is)

3 Likes
5

@pike I am aware how email works, I’ve set up many mail servers in my years in IT (I have to admit I was a bit triggered by your piping example - I am 20+ years in IT, I really don’t need plumbing examples). I don’t have an issue with my mail host polling it in less then 5 minutes periods (and if I had an issue with that, I would not be asking here).

@mrmarkuz thanks I will try that!

@mark_nl well this is good. I consider this a feature, I don’t want overlapping gets.

EDIT: Did the changes, I see the proper settings in getmail’s cron file. I will do the tests, but I guess this is it.

I wonder why NS doesn’t allow for lower settings in the interface (but default to 5).
If a host has an issue, then it won’t work anyway.

6

@NLS

Hi Nick

If you’ve been 20+ years in IT, good for you! But if you’re still running a mail server for no special reason with “pull”, you obviously haven’t kept up with best practices, if you’re still using a 1990’s method of handling mail.

I will admit to also having a client with NethServer with “pull”. The reason is he migrated from Windows SBS 2003 to NethSevrer in March 2020, just before corona lockdown. We do intend to move away from “pull”, but certain infrastructure needed to be in place first, like new firewalls, etc.
However, hardware delays, and all the other corona based issues delayed everything quite a bit. But best of all: my client is convinced he made the best move using NethServer instead of another Windows Server…
I’m using a 2min pull frequency here, as the Internet speed isn’t high enough, even with two providers and automatic failover. Besides which, the Internet hoster having the mailserver has worse uptime than we do!

Mail pulling also decreases the quality of Spam filtering, as the recieving server can’t follow the smtp exchange from the sender in full.

There are also other issues with pull, but that’s besides the point: Pull should NOT be used anymore without a good, valid reason…

My 2 cents
Andy

1 Like
7

Thanks man, I know the limitations and issues of using pull email and I can assure you there are specific reasons this happens like that in that specific domain.
Other domains I use, are using proper push or actually not use an intermediate host at all.

In any case the proposed edits above worked fine.

2 Likes