Expectation of when 2.1 will exit Beta?

jimp · Oct 1, 2012, 6:31 PM

There are some important things that need addressing, but they don't affect everyone.

For instance, there is an issue with NanoBSD on certain CF cards that causes a slow read-write to read-only transition. This needs fixing, but it only affects a portion of the user base. So that would hold up a release, but others may use it without issue at the same time.

We could probably at least put out a "real" beta image (BETA1, announced, with a set of images for download) with a list of known issues.

We still have to get 2.0.2 out as well…

maverick_slo · Oct 2, 2012, 6:16 AM

And while you mentioned 2.0.2, which mpd does it have, 5.5 or 5.6 ?
Thanks!

jimp · Oct 2, 2012, 2:44 PM

[2.0.2-RELEASE][root@pfs202-amd64]/root(7): mpd4 -v
Version 4.4.1 (root@base-8_1-amd64.builders.pfsense.org 21:23  6-Aug-2012)
[2.0.2-RELEASE][root@pfs202-amd64]/root(8): mpd5 -v
Version 5.6 (root@base-8_1-amd64.builders.pfsense.org 21:23  6-Aug-2012)

frater · Oct 2, 2012, 2:53 PM

@markuhde:

There's still a LOT of work that needs done. AFAIK, the traffic shaper doesn't work with VLAN's yet (that's kinda a big deal for some of us).

This is a big deal for me and to maybe more than just a few….
Can anyone confirm it doesn't work on the latest nightlies?
Where can I see the ticket for this?

xbipin · Oct 2, 2012, 3:35 PM

nanobsd ro to rw to ro is main issue, i tried the same on 3 CF cards, one MLC and 2 SLC based CF cards and its very very slow making changes when the CF is almost 80% full

xbipin · Oct 2, 2012, 3:37 PM

other things which i see need to be added is making upnp avoid feeding in rules which bypass limiters and this has been pending since a very long time now

jimp · Oct 2, 2012, 3:39 PM

Yes it is a big issue, it's not 100% related to the fullness of the drive but that could be partially to blame.

I have two cards here that are good examples:
1. Sandisk 4GB - freshly imaged, 4 seconds to switch at most
2. Kingston 4GB - freshly imaged, 30+ seconds to change

So there are multiple factors at play there.

xbipin · Oct 2, 2012, 3:47 PM

what i noticed was when drive was 45% full then it was quiet useable, once i installed freeradius2 to it, usage went 82% and then in captive portal i was adding pass through mac ids and each addition would take 45-60 seconds, then i removed freeradius2 and then tried and it became faster thats y i said about the fullness stuff

jimp · Oct 2, 2012, 3:48 PM

@xbipin:

other things which i see need to be added is making upnp avoid feeding in rules which bypass limiters and this has been pending since a very long time now

I'm not sure that's something we will see fixed in 2.1. Not sure though. It may have to wait for 2.2.

jimp · Oct 2, 2012, 3:49 PM

@xbipin:

what i noticed was when drive was 45% full then it was quiet useable, once i installed freeradius2 to it, usage went 82% and then in captive portal i was adding pass through mac ids and each addition would take 45-60 seconds, then i removed freeradius2 and then tried and it became faster thats y i said about the fullness stuff

Oh I don't doubt it - the filesystem code in question would likely behave that way. Though it may not be the fullness but the number of concurrent open files by FreeRADIUS. It's probably all part of the same issue there.

xbipin · Oct 2, 2012, 4:17 PM

has this issue come up with the new freebsd version or is it pfsense related?