- 1. e1000 w/ NAPI + SMP = 99% CPU utilization (score: 1)
- Author: inger@xxxxxxxx>
- Date: Mon, 07 Jun 2004 15:08:40 -0400
- With a stock 2.6.5 kernel, I'm building the e1000 driver as a module w/ NAPI turned on for an SMP host (Dell PowerEdge 1650 with 4 1Gb Intel NICs). ksoftirqd/0 is using 99% CPU utilization. However,
- /archives/netdev/2004-06/msg00162.html (7,636 bytes)
- 2. Re: e1000 w/ NAPI + SMP = 99% CPU utilization (score: 1)
- Author: jamal <hadi@xxxxxxxxxx>
- Date: Tue, 08 Jun 2004 13:34:52 +0100
- Chris Carpinello wrote: With a stock 2.6.5 kernel, I'm building the e1000 driver as a module w/ NAPI turned on for an SMP host (Dell PowerEdge 1650 with 4 1Gb Intel NICs). ksoftirqd/0 is using 99% CP
- /archives/netdev/2004-06/msg00197.html (8,986 bytes)
- 3. Re: e1000 w/ NAPI + SMP = 99% CPU utilization (score: 1)
- Author: r <shemminger@xxxxxxxx>
- Date: Tue, 08 Jun 2004 14:14:17 -0400
- Padraig wrote: At what packet rate does it go to 100%? I haven't narrowed down a threshold. tcpstat reports bps=202737465 on eth3. eth0 is a management interface (doesn't packet sniff). eth1 and eth2
- /archives/netdev/2004-06/msg00204.html (10,526 bytes)
- 4. Re: e1000 w/ NAPI + SMP = 99% CPU utilization (score: 1)
- Author: atiev <vkondra@xxxxxxx>
- Date: Wed, 9 Jun 2004 09:51:18 +0200
- Hello! Is seems like your network load @ ~202 Mbps gets you system into continuing polling as we see very few interrupts on your eth3. This means that rx_softirq reschedules itself do_softirq() kicks
- /archives/netdev/2004-06/msg00234.html (12,018 bytes)
- 5. Re: e1000 w/ NAPI + SMP = 99% CPU utilization (score: 1)
- Author: <cks@xxxxxxxxxxxxxxxx>
- Date: Wed, 09 Jun 2004 10:01:02 +0100
- Robert Olsson wrote: Chris Carpinello writes: Hello! Is seems like your network load @ ~202 Mbps gets you system into continuing polling as we see very few interrupts on your eth3. This means that rx
- /archives/netdev/2004-06/msg00236.html (9,334 bytes)
- 6. e1000 w/ NAPI + SMP = 99% CPU utilization (score: 1)
- Author: "Chris Carpinello" <chriscarpinello@xxxxxxxxxxx>
- Date: Mon, 07 Jun 2004 15:08:40 -0400
- With a stock 2.6.5 kernel, I'm building the e1000 driver as a module w/ NAPI turned on for an SMP host (Dell PowerEdge 1650 with 4 1Gb Intel NICs). ksoftirqd/0 is using 99% CPU utilization. However,
- /archives/netdev/2004-06/msg01046.html (7,525 bytes)
- 7. Re: e1000 w/ NAPI + SMP = 99% CPU utilization (score: 1)
- Author: P@xxxxxxxxxxxxxx
- Date: Tue, 08 Jun 2004 13:34:52 +0100
- Chris Carpinello wrote: With a stock 2.6.5 kernel, I'm building the e1000 driver as a module w/ NAPI turned on for an SMP host (Dell PowerEdge 1650 with 4 1Gb Intel NICs). ksoftirqd/0 is using 99% CP
- /archives/netdev/2004-06/msg01081.html (8,932 bytes)
- 8. Re: e1000 w/ NAPI + SMP = 99% CPU utilization (score: 1)
- Author: "Chris Carpinello" <chriscarpinello@xxxxxxxxxxx>
- Date: Tue, 08 Jun 2004 14:14:17 -0400
- I haven't narrowed down a threshold. tcpstat reports bps=202737465 on eth3. eth0 is a management interface (doesn't packet sniff). eth1 and eth2 are ifconfig'd down. Anyway it's not much to worry abo
- /archives/netdev/2004-06/msg01088.html (10,366 bytes)
- 9. Re: e1000 w/ NAPI + SMP = 99% CPU utilization (score: 1)
- Author: Robert Olsson <Robert.Olsson@xxxxxxxxxxx>
- Date: Wed, 9 Jun 2004 09:51:18 +0200
- Hello! Is seems like your network load @ ~202 Mbps gets you system into continuing polling as we see very few interrupts on your eth3. This means that rx_softirq reschedules itself do_softirq() kicks
- /archives/netdev/2004-06/msg01118.html (12,088 bytes)
- 10. Re: e1000 w/ NAPI + SMP = 99% CPU utilization (score: 1)
- Author: P@xxxxxxxxxxxxxx
- Date: Wed, 09 Jun 2004 10:01:02 +0100
- Hello! Is seems like your network load @ ~202 Mbps gets you system into continuing polling as we see very few interrupts on your eth3. This means that rx_softirq reschedules itself do_softirq() kicks
- /archives/netdev/2004-06/msg01120.html (9,512 bytes)
This search system is powered by
Namazu