Re: so much clock interrupts?!

2005-11-29 Thread cpghost
On Sat, Nov 26, 2005 at 10:27:42PM -0500, Chuck Swiger wrote: I'm also wondering wether 1000 Hz on a Soekris net4801 (Geode 266 MHz) won't be overkill. I'm planning to migrate some of them from 5.4 to 6.0, and doubting wether to change the new default to its more conservative previous setting

Re: so much clock interrupts?!

2005-11-26 Thread Chuck Swiger
cpghost wrote: On Fri, Nov 25, 2005 at 07:45:03AM +0100, Markus Trippelsdorf wrote: Yes, I guess it increases performance on a high throughput webserver or router that uses polling. But on the Desktop it only increases the overhead without any benefits at all. 2000 interrupts per second per

Re: so much clock interrupts?!

2005-11-25 Thread Kris Kennaway
On Fri, Nov 25, 2005 at 07:45:03AM +0100, Markus Trippelsdorf wrote: On Thu, Nov 24, 2005 at 03:43:59PM -0500, Kris Kennaway wrote: On Thu, Nov 24, 2005 at 07:40:40AM +, Markus Trippelsdorf wrote: On 2005-11-24, Wojciech Puchar [EMAIL PROTECTED] wrote: that's what i see in systat

Re: so much clock interrupts?!

2005-11-25 Thread cpghost
On Fri, Nov 25, 2005 at 07:45:03AM +0100, Markus Trippelsdorf wrote: Yes, I guess it increases performance on a high throughput webserver or router that uses polling. But on the Desktop it only increases the overhead without any benefits at all. 2000 interrupts per second per core for the

Re: so much clock interrupts?!

2005-11-25 Thread Wojciech Puchar
I'm also wondering wether 1000 Hz on a Soekris net4801 (Geode 266 MHz) won't be overkill. I'm planning to migrate some of them from 5.4 to 6.0, probably not. my 400Mhz PII works almost as fast with HZ=1000 and HZ=100 - tested today and doubting wether to change the new default to its more

Re: so much clock interrupts?!

2005-11-25 Thread RW
On Friday 25 November 2005 06:45, Markus Trippelsdorf wrote: 2000 interrupts per second per core for the timer is a ridiculous high number and I reduce it simply for aesthetic reasons. This may be a religious issue and everyone should use what he or she seems fitting. If you think in a

Re: so much clock interrupts?!

2005-11-24 Thread Wojciech Puchar
on FreeBSD 5.3 live CD i have 100 not 2000 times per second isn't it too much?! Yes it is. That's why I have kern.hz=100 in my /boot/loader.conf . -- do you know why they set it too high in 6.0? there must be a reason! ___

Re: so much clock interrupts?!

2005-11-24 Thread Wojciech Puchar
540 534 99 pdwak 2000 cpu0: time 16825 pdpgs 2000 cpu1: time on FreeBSD 5.3 live CD i have 100 not 2000 times per second isn't it too

Re: so much clock interrupts?!

2005-11-24 Thread Markus Trippelsdorf
On 2005-11-24, Wojciech Puchar [EMAIL PROTECTED] wrote: on FreeBSD 5.3 live CD i have 100 not 2000 times per second isn't it too much?! Yes it is. That's why I have kern.hz=100 in my /boot/loader.conf . -- do you know why they set it too high in 6.0? there must be a reason! Check:

Re: so much clock interrupts?!

2005-11-24 Thread Kris Kennaway
On Thu, Nov 24, 2005 at 07:40:40AM +, Markus Trippelsdorf wrote: On 2005-11-24, Wojciech Puchar [EMAIL PROTECTED] wrote: that's what i see in systat with FreeBSD 6.0/i386-SMP 540 534 99 pdwak 2000 cpu0: time

Re: so much clock interrupts?!

2005-11-24 Thread Kris Kennaway
On Thu, Nov 24, 2005 at 02:05:16PM +0100, Wojciech Puchar wrote: 540 534 99 pdwak 2000 cpu0: time 16825 pdpgs 2000 cpu1:

Re: so much clock interrupts?!

2005-11-24 Thread Wojciech Puchar
That's why I have kern.hz=100 in my /boot/loader.conf . Have you been able to measure a performance benefit from reducing it? Have you shown that reducing it does not cause your performance to *drop*? It was increased for a reason..it actually increases performance on some workloads. no i'm

Re: so much clock interrupts?!

2005-11-24 Thread Wojciech Puchar
No, it's just a consequence of HZ=1000 instead of HZ=100. I've so why 2000 not 1000? 2000 on each processor! Technical reasons..anyway, 2000 might look like a large number to you, but it's really not unless you're on a very slow machine (like a 486). i'm asking why it's 2000, not 1000.

Re: so much clock interrupts?!

2005-11-24 Thread Kris Kennaway
On Fri, Nov 25, 2005 at 01:09:08AM +0100, Wojciech Puchar wrote: No, it's just a consequence of HZ=1000 instead of HZ=100. I've so why 2000 not 1000? 2000 on each processor! Technical reasons..anyway, 2000 might look like a large number to you, but it's really not unless you're on a

Re: so much clock interrupts?!

2005-11-24 Thread Kris Kennaway
On Fri, Nov 25, 2005 at 01:08:42AM +0100, Wojciech Puchar wrote: That's why I have kern.hz=100 in my /boot/loader.conf . Have you been able to measure a performance benefit from reducing it? Have you shown that reducing it does not cause your performance to *drop*? It was increased for a

Re: so much clock interrupts?!

2005-11-24 Thread Markus Trippelsdorf
On Thu, Nov 24, 2005 at 03:43:59PM -0500, Kris Kennaway wrote: On Thu, Nov 24, 2005 at 07:40:40AM +, Markus Trippelsdorf wrote: On 2005-11-24, Wojciech Puchar [EMAIL PROTECTED] wrote: that's what i see in systat with FreeBSD 6.0/i386-SMP 540 534 99

Re: so much clock interrupts?!

2005-11-23 Thread Kris Kennaway
On Thu, Nov 24, 2005 at 01:35:41AM +0100, Wojciech Puchar wrote: that's what i see in systat with FreeBSD 6.0/i386-SMP 540 534 99 pdwak 2000 cpu0: time 16825 pdpgs 2000 cpu1:

Re: so much clock interrupts?!

2005-11-23 Thread Markus Trippelsdorf
On 2005-11-24, Wojciech Puchar [EMAIL PROTECTED] wrote: that's what i see in systat with FreeBSD 6.0/i386-SMP 540 534 99 pdwak 2000 cpu0: time 16825 pdpgs 2000 cpu1: time on FreeBSD