Since nv_sata2 and nge0 share the same interrupt vector, so when the interrupt happens, both the interrupt handler for nv_sata2 and nge0 will be called. You can setup your network on nge1 link to avoid this issue. If you must use nge0, you can check source code for nv_sata and nge to know what key function are called by interrupt handler after the basic device status checking and use dtrace to counter the call number of these key function.
On Tue, Feb 2, 2010 at 3:55 PM, Qihua Wu <dtrace...@gmail.com> wrote: > Yes, so if there share one interrupt vector, the number shown in intrstat > will be the same? Any way to know how many is from disk and how many is from > network? > > # echo ::interrupts -d | mdb -k > IRQ Vector IPL Bus Type CPU Share APIC/INT# Driver Name(s) > 4 0xb0 12 ISA Fixed 4 1 0x0/0x4 asy#0 > 9 0x81 9 PCI Fixed 1 1 0x0/0x9 acpi_wrapper_isr > 16 0x40 5 PCI Fixed 3 1 0x0/0x10 aac#0 > 20 0x41 5 PCI Fixed 5 2 0x0/0x14 nv_sata#0, ehci#0 > 21 0x42 5 PCI Fixed 6 2 0x0/0x15 nv_sata#1, ohci#0 > 22 0x60 6 PCI Fixed 7 2 0x0/0x16 nv_sata#2, nge#0 > 23 0x61 6 PCI Fixed 7 1 0x0/0x17 nge#1 > 44 0x63 6 PCI Fixed 8 2 0x1/0x14 nge#3, nv_sata#3 > 45 0x44 5 PCI Fixed 8 1 0x1/0x15 nv_sata#4 > 46 0x45 5 PCI Fixed 8 1 0x1/0x16 nv_sata#5 > 47 0x62 6 PCI Fixed 7 1 0x1/0x17 nge#2 > 48 0x82 7 MSI 2 1 - pcie_pci#2 > 49 0x83 7 MSI 2 1 - pcie_pci#2 > 160 0xa0 0 IPI ALL 0 - poke_cpu > 192 0xc0 13 IPI ALL 1 - xc_serv > 208 0xd0 14 IPI ALL 1 - kcpc_hw_overflow_intr > 209 0xd1 14 IPI ALL 1 - cbe_fire > 210 0xd3 14 IPI ALL 1 - cbe_fire > 240 0xe0 15 IPI ALL 1 - xc_serv > 241 0xe1 15 IPI ALL 1 - apic_error_intr > > > On Tue, Feb 2, 2010 at 3:29 PM, zhihui Chen <zhch...@gmail.com> wrote: >> >> You can use "echo ::interrupts | mdb -k" to check whether nge and >> nv_sata share the same interrupt vector. >> >> On Tue, Feb 2, 2010 at 2:59 PM, Qihua Wu <dtrace...@gmail.com> wrote: >> > For X4100 M2, we have no interrupt on disk, after upgraded to X4440, we >> > got >> > interrupt on disk, and what's weird is that the interrupt on network >> > card is >> > always equal to the number of interrupt on disk. I also used iosnoop the >> > check what is touching the disk, iosnoop shows sched touched the disk, >> > not >> > sure why. >> > >> > X4100 M2 >> > phxdb158$> intrstat -c 0 >> > >> > device | cpu0 %tim >> > -------------+--------------- >> > e1000g#1 | 0 0.0 >> > ehci#0 | 0 0.0 >> > nge#0 | 3352 2.1 >> > >> > device | cpu0 %tim >> > -------------+--------------- >> > e1000g#1 | 0 0.0 >> > ehci#0 | 0 0.0 >> > mpt#0 | 0 0.0 >> > nge#0 | 3201 2.0 >> > >> > device | cpu0 %tim >> > -------------+--------------- >> > e1000g#1 | 0 0.0 >> > ehci#0 | 0 0.0 >> > mpt#0 | 0 0.0 >> > nge#0 | 3458 2.2 >> > >> > device | cpu0 %tim >> > -------------+--------------- >> > e1000g#1 | 0 0.0 >> > ehci#0 | 0 0.0 >> > mpt#0 | 0 0.0 >> > nge#0 | 3045 1.9 >> > >> > device | cpu0 %tim >> > -------------+--------------- >> > e1000g#1 | 0 0.0 >> > ehci#0 | 0 0.0 >> > mpt#0 | 0 0.0 >> > nge#0 | 3025 1.8 >> > >> > device | cpu0 %tim >> > -------------+--------------- >> > e1000g#1 | 0 0.0 >> > ehci#0 | 0 0.0 >> > mpt#0 | 0 0.0 >> > nge#0 | 3206 1.9 >> > >> > X4440 >> > phxdb611$> intrstat -c 7 >> > >> > device | cpu7 %tim >> > -------------+--------------- >> > ehci#0 | 0 0.0 >> > nge#0 | 6983 2.9 >> > nge#1 | 3 0.0 >> > nv_sata#0 | 0 0.0 >> > nv_sata#2 | 6983 0.6 ====> disk interrupt, always equal to the >> > interrupt on nge#0, why this happens. >> > >> > device | cpu7 %tim >> > -------------+--------------- >> > ehci#0 | 0 0.0 >> > nge#0 | 6944 2.9 >> > nge#1 | 4 0.0 >> > nv_sata#0 | 0 0.0 >> > nv_sata#2 | 6944 0.6 >> > >> > device | cpu7 %tim >> > -------------+--------------- >> > ehci#0 | 0 0.0 >> > nge#0 | 7177 2.9 >> > nge#1 | 4 0.0 >> > nv_sata#0 | 0 0.0 >> > nv_sata#2 | 7177 0.6 >> > >> > device | cpu7 %tim >> > -------------+--------------- >> > ehci#0 | 0 0.0 >> > nge#0 | 6908 2.9 >> > nge#1 | 3 0.0 >> > nv_sata#0 | 0 0.0 >> > nv_sata#2 | 6908 0.6 >> > >> > device | cpu7 %tim >> > -------------+--------------- >> > ehci#0 | 0 0.0 >> > nge#0 | 6722 2.9 >> > nge#1 | 4 0.0 >> > nv_sata#0 | 0 0.0 >> > nv_sata#2 | 6722 0.6 >> > >> > device | cpu7 %tim >> > -------------+--------------- >> > aac#0 | 0 0.0 >> > ehci#0 | 0 0.0 >> > nge#0 | 6579 2.7 >> > nge#1 | 4 0.0 >> > nv_sata#0 | 0 0.0 >> > nv_sata#2 | 6579 0.6 >> > >> > phxdb611$> iosnoop -a >> > STRTIME DEVICE MAJ MIN UID PID D BLOCK >> > SIZE PATHNAME ARGS >> > 2010 Feb 1 23:47:59 sd1 31 70 0 3 W 2875520 >> > 8192 /var/adm/pacct fsflush\0 >> > 2010 Feb 1 23:47:59 sd1 31 69 0 3 W 11872 8192 >> > /tmp/hsperfdata_root/249 fsflush\0 >> > 2010 Feb 1 23:48:06 sd1 31 70 0 3 W 253224 1024 >> > /var/perfstats/vmstat.phxdb611.out_0201_230000 fsflush\0 >> > 2010 Feb 1 23:48:12 sd1 31 64 0 3 W 55376 >> > 15360 <none> fsflush\0 >> > 2010 Feb 1 23:48:12 sd1 31 70 0 3 W 29210 >> > 11264 <none> fsflush\0 >> > 2010 Feb 1 23:48:12 sd1 31 69 0 3 W 11520 >> > 14336 <none> fsflush\0 >> > 2010 Feb 1 23:48:12 sd1 31 68 0 3 W 105634 >> > 512 <none> fsflush\0 >> > 2010 Feb 1 23:48:14 sd1 31 70 0 3 W 252824 2048 >> > /var/perfstats/iostat.phxdb611.out_0201_230000 fsflush\0 >> > 2010 Feb 1 23:48:19 sd1 31 69 0 0 W 16 >> > 2048 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 69 0 0 W 48 >> > 8192 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 69 0 0 W 64 >> > 1024 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 69 0 0 W 73 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 69 0 0 W 1552 >> > 1024 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 69 0 0 W 1584 >> > 1024 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 16 >> > 2048 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 1260 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 295377 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 295395 >> > 1024 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 295403 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 295407 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 295414 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 295416 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 295429 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 590676 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 590695 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 689553 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 689603 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 689640 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 689681 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 689720 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 689734 >> > 1024 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 689737 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 689747 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 788232 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 49936 >> > 1024 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 R 886608 >> > 8192 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 295440 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 295487 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 295583 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 295625 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 295651 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 295699 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 296581 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 296586 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 296590 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 886286 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 886305 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 886310 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 886435 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 886495 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 886498 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 886501 >> > 1024 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 886506 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 886525 >> > 1024 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 886553 >> > 1536 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 886611 >> > 5632 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 886643 >> > 1024 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 49936 >> > 1024 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 296597 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 296600 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 296608 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 296618 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 296632 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 296637 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 296662 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 296753 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 492226 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 492359 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 492597 >> > 1536 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 492999 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 886658 >> > 4096 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 886667 >> > 2560 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 886672 >> > 1536 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 984676 >> > 1024 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 984882 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 984884 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 1082990 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 1083087 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 1083088 >> > 1024 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 1083098 >> > 1536 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 49936 >> > 1024 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 1082800 >> > 8192 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 1083104 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 1083110 >> > 512 <none> sched\0 >> > 2010 Feb 1 23:48:19 sd1 31 64 0 0 W 49936 >> > 1024 <none> sched\0 >> > 2010 Feb 1 23:48:21 sd1 31 70 0 3 W 253112 4096 >> > /var/perfstats/netstats-phxdb611-000051414.txt fsflush\0 >> > ^C >> > >> > _______________________________________________ >> > dtrace-discuss mailing list >> > dtrace-discuss@opensolaris.org >> > > > _______________________________________________ dtrace-discuss mailing list dtrace-discuss@opensolaris.org
