Andy Furniss wrote: > Patrick McHardy wrote: > >> Sorry for the repost Dave, I accidentally used the @oss.sgi.com address >> in my previous posting. >> >> --------- >> >> It seems to be a common mistake to use jiffies as clocksource, >> which gives very bad results in most cases. This patch changes >> the default to gettimeofday. > > > What do you think about making HTB hysteresis 0 and possibly set HZ to > 1000 when HTB is selected?
I'm not qualified to judge about HTB hysteresis, but we can't change HZ, most people just enable everything which would mean basically everybody would be using HZ=1000 again. > By default now I think jitter on htb is 8ms on 100mbit it's 1ms with the > tweaks (depending on there being 1 class). > > Irrespective of Hz, at low speed - say 128kbit the default setting of > HYSTERSIS 1 means that bulk packets are sent in pairs unneccerily adding > 90ms to jitter. > > HTB default qlen is that of the device or 3 if it is less. Maybe it > should be increased 30? 3 is pretty useless for thoughput, but means > that people who don't specify/get default lengths on sub queues get a > half working setup. That only affects the "direct queue" for unclassified packets, doesn't it? - To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
