OPNsense Forum

English Forums => General Discussion => Topic started by: oscarr on August 05, 2020, 12:31:03 am

Title: How the pipes actually work?
Post by: oscarr on August 05, 2020, 12:31:03 am

Hi,
I saw lots of tutorials about QOS with pipes and rules, but none of them explained how the pipes will act in certain situations.

In FreshTomato QOS is also based on kind of pipes, but there its clear how it works.
In there pipes are set for each service. You can set minimal guaranteed bandwidth for each service, but maximum is not set, so if other services are not using their bandwidth, other services can use even 100%.

I wonder how it works in OPNsense? I can set proportions between pipes, but how it is when one pipe is empty - not in use ... is its bandwidth still reserved or can be used by other service temporarily?

Thanks
Al

Title: Re: How the pipes actually work?
Post by: oscarr on August 09, 2020, 08:30:00 pm

OK, I think I have found documentation explaining how to do what I'm looking for.

https://docs.opnsense.org/manual/how-tos/shaper_prioritize_using_queues.html (https://docs.opnsense.org/manual/how-tos/shaper_prioritize_using_queues.html)

I didn't recognise on first look because example is a bit confusing. In example weight 9 means 9MBit which suggest that weight is specified in megabytes, but in fact weight is stated as percentage of the pipe bandwidth.

So now I have pipes, I made the queues, but stucked on rules.
Thing is, that I'm afraid I'm confusing what is source what is destination.
What is source port and what is destination?