does any body ever met this strange problem?
hi, does any body ever met this strange problem?
--linux1---------linux2--------linux3--
eth0 eth1
my purpose is :
make the packets from linux1 to linux3
to be marked with a proper dscp value when
they gose through linux2.
i do the following at eth0 on linux2:
tc qdisc add dev eth0 handle ffff: ingress
tc filter add dev eth0 parent ffff: protocol ip prio 4 u32 \
match ip dst 10.0.2.18 flowid :1
and at eth1 on linux2 :
tc qdisc add dev eth1 handle 1:0 root dsmark indices 64 tc class change dev eth1 classid 1:1 dsmark mask 0x3 value 0x88 tc filter add dev eth1 parent 1:0 protocol all prio 4 \
handle 1 tcindex classid 1:1
now i ping linux3(10.0.2.18) from linux1 , and i can see the dscp value 0x22 catched at linux3. everything looks perfect now :)
But , when i add the following line at eth1 on linux2 the problem arised :
tc qdisc add dev eth1 parent 1:0 handle 2:0 dsmark \
indices 64 set_tc_index
tc filter add dev eth1 parent 2:0 protocol all prio \
1 tcindex mask 0xfc shift 2 pass_on
tc filter add dev eth1 parent 2:0 prio 1 handle 0x22\
tcindex classid 2:111
tc qdisc add dev eth1 parent 2:0 handle 3:0 cbq bandwidth \
10Mbit cell 8 avpkt 1000 mpu 64
tc filter add dev eth1 parent 3:0 protocol all prio 1 \
tcindex mask 0xf0 shift 4 pass_ontc class add dev eth1 parent 3:0 classid 3:99 cbq bandwidth \
10Mbit rate 9Mbit avpkt 1000 prio 2 \
allot 1514 weight 9 maxburst 21 defmap 0
tc filter add dev eth1 parent 3:0 prio 1 handle 1 tcindex \
classid 3:99
when i use 'tc -s class ls dev eth1'
i can see the packet from linux1 to linux3 is directed into class 3:99 , BUT the packet catched at linux3 have a dscp value of ZERO.
where has the dscp 0x22 gone?
_________________________________________________________________
Choose an Internet access plan right for you -- try MSN!
http://resourcecenter.msn.com/access/plans/default.asp
_______________________________________________
LARTC mailing list / [email protected]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
Quoted 1 line · Folke Aeon
hi, does any body ever met this strange problem?
No, I haven't, because I usually don't nest dsmark ;-)
dsmark changes the DS field according to the value in skb->tc_index when the packet _leaves_ the qdisc, and your inner dsmark changes skb->tc_index.
I suggest you have a look at figures 6 and 8 (the latter is on page
6) of ftp://icaftp.epfl.ch/pub/linux/diffserv/misc/dsid-01.ps.gz
- Werner
Signature · 6 lines
_________________________________________________________________________ / Werner Almesberger, Buenos Aires, Argentina [email protected] / /_http://www.almesberger.net/____________________________________________/ _______________________________________________ LARTC mailing list / [email protected] http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
Thank Werner,
that's helps alot :)
i managed to figure out that the following does the trick for me :
tc qdisc add dev eth1 handle 1:0 root dsmark indices 64 tc class change dev eth1 classid 1:11 dsmark mask 0x3 value 0x28 tc class change dev eth1 classid 1:21 dsmark mask 0x3 value 0x48 tc filter add dev eth1 parent 1:0 protocol all prio 1 handle 2 \
tcindex classid 1:11
tc filter add dev eth1 parent 1:0 protocol all prio 1 handle 1 \
tcindex classid 1:21
#tc qdisc add dev eth1 parent 1:0 handle 2:0 dsmark indices 64 set_tc_index #tc filter add dev eth1 parent 2:0 protocol all prio 1 tcindex mask 0xfc shift 2 pass_on
#tc filter add dev eth1 parent 2:0 prio 1 handle 0x22 tcindex classid 2:111 #tc filter add dev eth1 parent 2:0 prio 1 handle 0x12 tcindex classid 2:191
tc qdisc add dev eth1 parent 1:0 handle 2:0 cbq bandwidth \
10Mbit cell 8 avpkt 1000 mpu 64
tc filter add dev eth1 parent 2:0 protocol all prio 1 \
tcindex mask 0xf0 shift 4 pass_ontc class add dev eth1 parent 2:0 classid 2:99 cbq bandwidth \
10Mbit rate 9Mbit avpkt 1000 prio 2 \
allot 1514 weight 9 maxburst 21 defmap 0
tc class add dev eth1 parent 2:0 classid 2:44 cbq bandwidth \
10Mbit rate 1Mbit avpkt 1000 prio 2 \
allot 1514 weight 9 maxburst 21 defmap 0
tc filter add dev eth1 parent 2:0 prio 1 handle 1 tcindex \
classid 2:99
tc filter add dev eth1 parent 2:0 prio 1 handle 2 tcindex \
classid 2:44tc qdisc add dev eth1 parent 2:99 handle 4:0 pfifo limit 10 tc qdisc add dev eth1 parent 2:44 handle 5:0 pfifo limit 10
note that after i delete the lines begin with notation mark "#",
all things seems to be back to normal. :)
will someone give some comment on this ?
Quoted 17 lines
From: Werner Almesberger <[email protected]>
To: Folke Aeon <[email protected]>
CC: Subject: Re: [LARTC] does any body ever met this strange problem?
Date: Thu, 24 Oct 2002 03:19:51 -0300Quoted 1 line · Folke Aeon
hi, does any body ever met this strange problem?
No, I haven't, because I usually don't nest dsmark ;-)
dsmark changes the DS field according to the value in
skb->tc_indexwhen the packet _leaves_ the qdisc, and your inner dsmark changesskb->tc_index.I suggest you have a look at figures 6 and 8 (the latter is on page
6) of ftp://icaftp.epfl.ch/pub/linux/diffserv/misc/dsid-01.ps.gz
Quoted 8 lines
- Werner
--
_________________________________________________________________________
/ Werner Almesberger, Buenos Aires, Argentina [email protected]
/ /_http://www.almesberger.net/____________________________________________/
_________________________________________________________________
Unlimited Internet access -- and 2 months free! Try MSN.
http://resourcecenter.msn.com/access/plans/2monthsfree.asp
_______________________________________________
LARTC mailing list / [email protected]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/