David S. Miller wrote: > From: Ben Greear <greearb@candelatech.com> > Date: Wed, 28 Aug 2002 00:49:57 -0700 > > At the time I wrote the vlan code, ... > > Ben I basically rewrote nearly the entirety of your VLAN > stack. Excellent. One of us buggered this up, try: vconfig add eth0 0 vconfig rem eth0 0 > > Next, the lookups happen during ifconfig, and we found this > to be a bug in how ifconfig itself worked. ifconfig is still much slower and 'ip'. I'm not sure if that can be helped or not though. At any rate, using RH 7.3 ifconfig and the latest pre4 kernel, I see these numbers on my 2Ghz dual-proc athlon: ifconfig -a for 4000 vlans: real 44.38 user 2.76 sys 41.62 ip addr show for 4000 vlans: real 5.57 user 1.72 sys 3.86 So, it's not as bad as it used to be, but ifconfig -a is still pretty damn slow. It's all in the kernel too, though ifconfig is probably abusing the kernel in some way that 'ip' does not. > Just like the first time you brought this up, and now for the > umpteenth time, I totally disagree. Looking up by index is > totally slow path and therefore there is no need to complicate > it. So, for the record, is 44 seconds on a high-end machine acceptable for ifconfig -a on 4000 interfaces? It doesn't bother me enough to re-write a patch that will never be accepted, but if it were my code, I'd change it to make it faster :) Take it easy, Ben -- Ben Greear <greearb@candelatech.com> <Ben_Greear AT excite.com> President of Candela Technologies Inc http://www.candelatech.com ScryMUD: http://scry.wanfear.com http://scry.wanfear.com/~greear - : send the line "unsubscribe linux-net" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
