* [dtaht/ceropackages-3.10] 27eed1: Initial changes to squash inbound diffserv and hav...
@ 2014-04-09 21:29 GitHub
0 siblings, 0 replies; only message in thread
From: GitHub @ 2014-04-09 21:29 UTC (permalink / raw)
To: cerowrt-commits
[-- Attachment #1: Type: text/plain, Size: 1439 bytes --]
Branch: refs/heads/master
Home: https://github.com/dtaht/ceropackages-3.10
Commit: 27eed160a67700caae85a4c8b3fff0eaa990cd27
https://github.com/dtaht/ceropackages-3.10/commit/27eed160a67700caae85a4c8b3fff0eaa990cd27
Author: Dave Taht <dave.taht@bufferbloat.net>
Date: 2014-04-09 (Wed, 09 Apr 2014)
Changed paths:
M net/sqm-scripts/files/usr/lib/sqm/functions.sh
M net/sqm-scripts/files/usr/lib/sqm/run.sh
M net/sqm-scripts/files/usr/lib/sqm/simple.qos
Log Message:
-----------
Initial changes to squash inbound diffserv and have simpler inbound
1) By default dscp values cannot be trusted e2e as multiple
providers change them in transit. So the correct "Default" for sqm
should be to reset all inbound diffserv marked values to
BE, and then reclassify.
2) It looks like the inbound rules for simple.qos were always
broken - see iptables -x -v -L for more details
3) IF you are not going to pay attention to diffserv,
it makes no sense to have a 3 tier system for inbound,
so this grabs simplest.qos's scheme and uses that if SQUASH_DSCP
is set.
4) Haven't tried updating the gui yet.
5) Bug:
It still looks like the tc based rules for diffserv *never* worked,
and I'm not sure which would be faster - iptables or diffserv
for outbound, so I wanted to fiddle with diffserv on that
and ran out of time - the code needs to use different priorities
((prio X)+1)) to work right in functions.sh
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2014-04-09 21:29 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-04-09 21:29 [dtaht/ceropackages-3.10] 27eed1: Initial changes to squash inbound diffserv and hav GitHub
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox