Wiki update for traffic shaper?
-
So, anyone up for the task? Still reading through threads, and I'm not sure what information is up to date, and some of the posts are conflicting, and a few others are a teensy bit in-depth and tbh, way over my head.
-
+1
I am kind of lost too -
Me too, I know that it is so complex but for now I am completly out of order…
DuroM
-
Well, the main issues seem to be with what rules match what traffic. Some things that would seem to make sense don't quite work, as stated in some of the 'way over my head' threads, and there seems to be conflicting information even in the threads that seem to have everything in order. And again there is the issue of which thread has the correct information. It seems as though very few people can figure out the new shaper.
-
There are still people having issues with the traffic shaper, and it seems to be at least partly due to no-one understanding the new shaper, conflicting information from different posts.
Anyone up to the task of updating the wiki? If I had all the information myself, I'd make an account on the wiki and update it – but that would be like the blind leading the blind.
-
Devs have to make it work first. Momentarily it does not work so it is quite pointless to update the docs.
-
While I agree with the statement on general principle, there is also the issue of testing the shaper so that we can get pfSense to stable, which would require testing all of the features. With no proper documentation, so that people can test it, it is harder for people to properly test this aspect of the firewall.
While perhaps a full-blown Wiki article may be questionable, perhaps a beta-testing wiki page, or something else to that effect would be appropriate? -
So is there anyone that can even put up an un-official guide of sorts for the traffic shaper? Noticing lots of posts lately about the shaper, and I think it would be good to have an up to date and accurate guide for the shaper, for both usability and testing. It is really hard to help test the shaper and get bugs ironed out when there isn't any place to help make heads or tails of it.
I understand that the shaper may be partially or moreso broken, but with RC2 coming along as it has, and hopefully a new RC coming soonish, figure that the shaper could use some serious looking into.
-
So is there anyone that can even put up an un-official guide of sorts for the traffic shaper? Noticing lots of posts lately about the shaper, and I think it would be good to have an up to date and accurate guide for the shaper, for both usability and testing. It is really hard to help test the shaper and get bugs ironed out when there isn't any place to help
The wiki didn't have the information I was looking for either so I decided to dig into it a bit and write a basic walk through of the shaper.
Unofficial guide to traffic shaping on pfSense 2.0
Comments, corrections, or additions are welcome.
-
That's really nice to have an unofficial guide.
I think the chicken and egg issue is facing shaping. It doesn't work. noone knows how to use it. While you can do some shaping functions, the wizards and internal gui leads one to believe they can use use for certain "well known" capabilities that were used in 1.2.x, which is misleading, because a lot of these don't work.
What is missing is a statement from the developers of what will be supported/fixed for 2.0 and what will be abandoned regarding shaping.
For example, shaping by ports is broken in the wizard and in general. L7 shaping works (supposedly) but only if the l7 filters are an exact match. Further, there is a resource need for any l7 shaping using DPI, and that has never been discussed or disclosed at to how many resources should be allocated and whether or not it can work reliably on a nano device (where resources are a bit more scarce).
I don't see shaping getting any attention until the devs decide to fix it or address it. They've been quiet when asked bluntly, so one can only assume they are not willing to discuss it in lieu of other priorities or because it is hopelessly broken. There are a lot of us that can't consider 2.x until this is addressed though, and I'm sure they are aware of it (somehow).