linux/net/sched
Paolo Valente 87f40dd6ce pkt_sched: sch_qfq: remove a source of high packet delay/jitter
QFQ+ inherits from QFQ a design choice that may cause a high packet
delay/jitter and a severe short-term unfairness. As QFQ, QFQ+ uses a
special quantity, the system virtual time, to track the service
provided by the ideal system it approximates. When a packet is
dequeued, this quantity must be incremented by the size of the packet,
divided by the sum of the weights of the aggregates waiting to be
served. Tracking this sum correctly is a non-trivial task, because, to
preserve tight service guarantees, the decrement of this sum must be
delayed in a special way [1]: this sum can be decremented only after
that its value would decrease also in the ideal system approximated by
QFQ+. For efficiency, QFQ+ keeps track only of the 'instantaneous'
weight sum, increased and decreased immediately as the weight of an
aggregate changes, and as an aggregate is created or destroyed (which,
in its turn, happens as a consequence of some class being
created/destroyed/changed). However, to avoid the problems caused to
service guarantees by these immediate decreases, QFQ+ increments the
system virtual time using the maximum value allowed for the weight
sum, 2^10, in place of the dynamic, instantaneous value. The
instantaneous value of the weight sum is used only to check whether a
request of weight increase or a class creation can be satisfied.

Unfortunately, the problems caused by this choice are worse than the
temporary degradation of the service guarantees that may occur, when a
class is changed or destroyed, if the instantaneous value of the
weight sum was used to update the system virtual time. In fact, the
fraction of the link bandwidth guaranteed by QFQ+ to each aggregate is
equal to the ratio between the weight of the aggregate and the sum of
the weights of the competing aggregates. The packet delay guaranteed
to the aggregate is instead inversely proportional to the guaranteed
bandwidth. By using the maximum possible value, and not the actual
value of the weight sum, QFQ+ provides each aggregate with the worst
possible service guarantees, and not with service guarantees related
to the actual set of competing aggregates. To see the consequences of
this fact, consider the following simple example.

Suppose that only the following aggregates are backlogged, i.e., that
only the classes in the following aggregates have packets to transmit:
one aggregate with weight 10, say A, and ten aggregates with weight 1,
say B1, B2, ..., B10. In particular, suppose that these aggregates are
always backlogged. Given the weight distribution, the smoothest and
fairest service order would be:
A B1 A B2 A B3 A B4 A B5 A B6 A B7 A B8 A B9 A B10 A B1 A B2 ...

QFQ+ would provide exactly this optimal service if it used the actual
value for the weight sum instead of the maximum possible value, i.e.,
11 instead of 2^10. In contrast, since QFQ+ uses the latter value, it
serves aggregates as follows (easy to prove and to reproduce
experimentally):
A B1 B2 B3 B4 B5 B6 B7 B8 B9 B10 A A A A A A A A A A B1 B2 ... B10 A A ...

By replacing 10 with N in the above example, and by increasing N, one
can increase at will the maximum packet delay and the jitter
experienced by the classes in aggregate A.

This patch addresses this issue by just using the above
'instantaneous' value of the weight sum, instead of the maximum
possible value, when updating the system virtual time.  After the
instantaneous weight sum is decreased, QFQ+ may deviate from the ideal
service for a time interval in the order of the time to serve one
maximum-size packet for each backlogged class. The worst-case extent
of the deviation exhibited by QFQ+ during this time interval [1] is
basically the same as of the deviation described above (but, without
this patch, QFQ+ suffers from such a deviation all the time). Finally,
this patch modifies the comment to the function qfq_slot_insert, to
make it coherent with the fact that the weight sum used by QFQ+ can
now be lower than the maximum possible value.

[1] P. Valente, "Extending WF2Q+ to support a dynamic traffic mix",
Proceedings of AAA-IDEA'05, June 2005.

Signed-off-by: Paolo Valente <paolo.valente@unimore.it>
Signed-off-by: David S. Miller <davem@davemloft.net>
2013-07-18 13:02:00 -07:00
..
Kconfig net: sched: enable CAN Identifier to be build into kernel 2012-11-25 16:06:06 -05:00
Makefile net: sched: add ipset ematch 2012-07-12 07:54:46 -07:00
act_api.c rtnetlink: Remove passing of attributes into rtnl_doit functions 2013-03-22 10:31:16 -04:00
act_csum.c act_csum: fix possible use after free 2013-04-12 15:25:41 -04:00
act_gact.c pkt_sched: namespace aware act_mirred 2013-01-14 15:09:36 -05:00
act_ipt.c net_sched: act_ipt forward compat with xtables 2013-05-01 13:19:19 -04:00
act_mirred.c net: pass info struct via netdevice notifier 2013-05-28 13:11:01 -07:00
act_nat.c pkt_sched: namespace aware act_mirred 2013-01-14 15:09:36 -05:00
act_pedit.c net: Add skb_unclone() helper function. 2013-02-15 15:10:37 -05:00
act_police.c net_sched: restore "overhead xxx" handling 2013-06-02 22:22:35 -07:00
act_simple.c pkt_sched: namespace aware act_mirred 2013-01-14 15:09:36 -05:00
act_skbedit.c pkt_sched: namespace aware act_mirred 2013-01-14 15:09:36 -05:00
cls_api.c net_cls: remove duplicated include from cls_api.c 2013-04-07 17:12:01 -04:00
cls_basic.c pkt_sched: namespace aware act_mirred 2013-01-14 15:09:36 -05:00
cls_cgroup.c pkt_sched: namespace aware act_mirred 2013-01-14 15:09:36 -05:00
cls_flow.c netlink: rename ssk to sk in struct netlink_skb_params 2013-04-19 14:57:56 -04:00
cls_fw.c pkt_sched: fix error return code in fw_change_attrs() 2013-04-19 17:34:53 -04:00
cls_route.c pkt_sched: namespace aware act_mirred 2013-01-14 15:09:36 -05:00
cls_rsvp.c
cls_rsvp.h pkt_sched: namespace aware act_mirred 2013-01-14 15:09:36 -05:00
cls_rsvp6.c
cls_tcindex.c pkt_sched: namespace aware act_mirred 2013-01-14 15:09:36 -05:00
cls_u32.c pkt_sched: namespace aware act_mirred 2013-01-14 15:09:36 -05:00
em_canid.c net: em_canid: Ematch rule to match CAN frames according to their identifiers 2012-07-04 13:07:05 +02:00
em_cmp.c net_sched: cleanups 2011-01-19 23:31:12 -08:00
em_ipset.c netfilter: ipset: Introduce extensions to elements in the core 2013-04-29 20:08:54 +02:00
em_meta.c net: use a per task frag allocator 2012-09-24 16:31:37 -04:00
em_nbyte.c net_sched: cleanups 2011-01-19 23:31:12 -08:00
em_text.c net_sched: cleanups 2011-01-19 23:31:12 -08:00
em_u32.c net_sched: cleanups 2011-01-19 23:31:12 -08:00
ematch.c net: Convert net_ratelimit uses to net_<level>_ratelimited 2012-05-15 13:45:03 -04:00
sch_api.c net_sched: qdisc_get_rtab() must check data[] array 2013-06-07 15:24:04 -07:00
sch_atm.c sch_atm.c: get rid of poinless extern 2012-06-01 10:37:18 -04:00
sch_blackhole.c
sch_cbq.c net_sched: add 64bit rate estimators 2013-06-11 02:51:03 -07:00
sch_choke.c net/sched: rename random32() to prandom_u32() 2013-04-29 18:28:43 -07:00
sch_codel.c fq_codel: should use qdisc backlog as threshold 2012-05-16 15:30:26 -04:00
sch_drr.c net_sched: add 64bit rate estimators 2013-06-11 02:51:03 -07:00
sch_dsmark.c net: sched: factorize code (qdisc_drop()) 2012-05-04 11:50:05 -04:00
sch_fifo.c pkt_sched: Stop using NLA_PUT*(). 2012-04-01 18:11:37 -04:00
sch_fq_codel.c net: fq_codel: Fix off-by-one error 2013-03-29 15:32:23 -04:00
sch_generic.c net_sched: psched_ratecfg_precompute() improvements 2013-06-11 22:39:47 -07:00
sch_gred.c net_sched: gred: actually perform idling in WRED mode 2012-09-13 16:10:13 -04:00
sch_hfsc.c net_sched: add 64bit rate estimators 2013-06-11 02:51:03 -07:00
sch_htb.c htb: refactor struct htb_sched fields for performance 2013-06-19 23:06:52 -07:00
sch_ingress.c net_sched: factorize qdisc stats handling 2011-01-10 16:07:54 -08:00
sch_mq.c pkt_sched: avoid requeues if possible 2012-12-12 00:16:47 -05:00
sch_mqprio.c pkt_sched: avoid requeues if possible 2012-12-12 00:16:47 -05:00
sch_multiq.c pkt_sched: Stop using NLA_PUT*(). 2012-04-01 18:11:37 -04:00
sch_netem.c netem: fix possible NULL deref in netem_dequeue() 2013-07-03 16:52:10 -07:00
sch_plug.c net_sched: sch_plug: plug_qdisc_ops is static 2012-02-13 16:04:40 -05:00
sch_prio.c pkt_sched: Stop using NLA_PUT*(). 2012-04-01 18:11:37 -04:00
sch_qfq.c pkt_sched: sch_qfq: remove a source of high packet delay/jitter 2013-07-18 13:02:00 -07:00
sch_red.c pkt_sched: Stop using NLA_PUT*(). 2012-04-01 18:11:37 -04:00
sch_sfb.c sch_sfb: Fix missing NULL check 2012-07-12 08:33:18 -07:00
sch_sfq.c pkt_sched: Stop using NLA_PUT*(). 2012-04-01 18:11:37 -04:00
sch_tbf.c Merge git://git.kernel.org/pub/scm/linux/kernel/git/davem/net 2013-06-05 16:37:30 -07:00
sch_teql.c sch_teql: Convert over to dev_neigh_lookup_skb(). 2012-07-05 01:09:06 -07:00