IRQ process using 60% CPU
IRQ process using 60% CPU
Is that normal for accel-ppp?
- Attachments
-
- zombie.png (53.59 KiB) Viewed 2701 times
Last edited by hugleo on 03 Feb 2017, 18:06, edited 1 time in total.
Re: Zombie process using 60% CPU
Most of times I see ksoftirqd/0 with 90%. Maybe it should be balancend between all cores ksoftirqd/1, ksoftirqd/2...
Code: Select all
CPU0 CPU1 CPU2 CPU3 CPU4 CPU5 CPU6 CPU7
0: 338093413 0 0 0 0 0 0 0 IO-APIC-edge timer
1: 5 0 0 0 0 0 0 0 IO-APIC-edge i8042
8: 1 0 0 0 0 0 0 0 IO-APIC-edge rtc0
9: 0 0 0 0 0 0 0 0 IO-APIC-fasteoi acpi
14: 0 0 0 0 0 0 0 0 IO-APIC-edge ata_piix
15: 0 0 0 0 0 0 0 0 IO-APIC-edge ata_piix
16: 43 0 0 0 0 0 0 0 IO-APIC-fasteoi uhci_hcd:usb3
18: 2 0 0 0 0 0 0 0 IO-APIC-fasteoi ehci_hcd:usb1, uhci_hcd:usb6
19: 0 0 0 0 0 0 0 0 IO-APIC-fasteoi uhci_hcd:usb5, ata_piix
23: 0 0 0 0 0 0 0 0 IO-APIC-fasteoi ehci_hcd:usb2, uhci_hcd:usb4
66: 108272 0 0 0 0 0 0 0 PCI-MSI-edge hpsa0
67: 323 330783 0 0 0 0 0 0 PCI-MSI-edge hpsa0
68: 115 0 371031 0 0 0 0 0 PCI-MSI-edge hpsa0
69: 64 0 0 369089 0 0 0 0 PCI-MSI-edge hpsa0
70: 179 0 0 0 20790 0 0 0 PCI-MSI-edge hpsa0
71: 85 0 0 0 0 65244 0 0 PCI-MSI-edge hpsa0
72: 249 0 0 0 0 0 61309 0 PCI-MSI-edge hpsa0
73: 242 0 0 0 0 0 0 53981 PCI-MSI-edge hpsa0
74: 0 0 0 0 0 0 0 0 PCI-MSI-edge eth0
75: 6 0 20 0 903535 2705 0 0 PCI-MSI-edge eth0-TxRx-0
76: 6 0 903535 20 0 2705 0 0 PCI-MSI-edge eth0-TxRx-1
77: 903561 0 0 0 0 2705 0 0 PCI-MSI-edge eth0-TxRx-2
78: 6 0 0 0 0 2725 903535 0 PCI-MSI-edge eth0-TxRx-3
79: 6 0 0 0 903535 2705 20 0 PCI-MSI-edge eth0-TxRx-4
80: 6 0 903535 0 0 2705 0 20 PCI-MSI-edge eth0-TxRx-5
81: 86 0 0 0 20 2705 0 903455 PCI-MSI-edge eth0-TxRx-6
82: 6 20 0 0 0 2705 903535 0 PCI-MSI-edge eth0-TxRx-7
83: 0 0 0 0 0 0 0 0 PCI-MSI-edge eth1
84: 7 0 903535 20 0 2705 0 0 PCI-MSI-edge eth1-TxRx-0
85: 107 0 0 0 0 2705 0 903455 PCI-MSI-edge eth1-TxRx-1
86: 7 0 0 0 0 2725 903535 0 PCI-MSI-edge eth1-TxRx-2
87: 7 0 0 0 903535 2705 20 0 PCI-MSI-edge eth1-TxRx-3
88: 7 0 903535 0 0 2705 0 20 PCI-MSI-edge eth1-TxRx-4
89: 87 0 0 0 20 2705 0 903455 PCI-MSI-edge eth1-TxRx-5
90: 7 20 0 0 0 2705 903535 0 PCI-MSI-edge eth1-TxRx-6
91: 7 0 20 0 903535 2705 0 0 PCI-MSI-edge eth1-TxRx-7
92: 1563132111 0 0 0 0 0 0 0 PCI-MSI-edge eth2-TxRx-0
93: 3247 398335202 0 0 0 0 0 0 PCI-MSI-edge eth2-TxRx-1
94: 4681 0 411653663 0 0 0 0 0 PCI-MSI-edge eth2-TxRx-2
95: 6756 0 0 454354532 0 0 0 0 PCI-MSI-edge eth2-TxRx-3
96: 5505 0 0 0 391022444 0 0 0 PCI-MSI-edge eth2-TxRx-4
97: 4963 0 0 0 0 353505547 0 0 PCI-MSI-edge eth2-TxRx-5
98: 3434 0 0 0 0 0 396123077 0 PCI-MSI-edge eth2-TxRx-6
99: 5179 0 0 0 0 0 0 363277592 PCI-MSI-edge eth2-TxRx-7
100: 3 0 177365 1 0 337 0 0 PCI-MSI-edge eth2
101: 2154369752 0 0 0 0 0 0 0 PCI-MSI-edge eth3-TxRx-0
102: 11 4161200363 0 0 0 0 0 0 PCI-MSI-edge eth3-TxRx-1
103: 63 0 4263319316 0 0 0 0 0 PCI-MSI-edge eth3-TxRx-2
104: 11 0 0 4246099053 0 0 0 0 PCI-MSI-edge eth3-TxRx-3
105: 50 0 0 0 3702980025 0 0 0 PCI-MSI-edge eth3-TxRx-4
106: 43 0 0 0 0 3764516372 0 0 PCI-MSI-edge eth3-TxRx-5
107: 59 0 0 0 0 0 3834224668 0 PCI-MSI-edge eth3-TxRx-6
108: 27 0 0 0 0 0 0 3762967399 PCI-MSI-edge eth3-TxRx-7
109: 3 0 0 0 0 0 0 4 PCI-MSI-edge eth3
NMI: 1648058 299857 307650 294591 524422 396746 387824 440960 Non-maskable interrupts
LOC: 3621212573 1878380193 1927318079 1959902316 1388530460 1495553213 1493434689 1471413791 Local timer interrupts
SPU: 0 0 0 0 0 0 0 0 Spurious interrupts
PMI: 1648058 299857 307650 294591 524422 396746 387824 440960 Performance monitoring interrupts
IWI: 3 0 0 0 0 0 3 0 IRQ work interrupts
RTR: 7 0 0 0 0 0 0 0 APIC ICR read retries
RES: 54309538 75220665 33431078 40677661 3326965 8127609 8303768 8190846 Rescheduling interrupts
CAL: 4294922491 4294893716 4294900660 4294895403 4294944733 4294933232 4294931023 4294927670 Function call interrupts
TLB: 2776300 2302116 2319603 2444776 640386 1092192 1144436 1165548 TLB shootdowns
TRM: 0 0 0 0 0 0 0 0 Thermal event interrupts
THR: 0 0 0 0 0 0 0 0 Threshold APIC interrupts
MCE: 0 0 0 0 0 0 0 0 Machine check exceptions
MCP: 6043 6043 6043 6043 6043 6043 6043 6043 Machine check polls
HYP: 0 0 0 0 0 0 0 0 Hypervisor callback interrupts
ERR: 0
MIS: 0
Re: Zombie process using 60% CPU
show perf top.
You use hyper threading?
You use hyper threading?
Re: Zombie process using 60% CPU
Using ht
dmidecode | grep Count
Core Count: 4
Thread Count: 8
dmidecode | grep Count
Core Count: 4
Thread Count: 8
Code: Select all
Samples: 357K of event 'cycles', Event count (approx.): 40901450135
13.11% [kernel] [k] pppoe_rcv
7.49% [kernel] [k] ipt_do_table
3.72% [kernel] [k] _raw_spin_lock
2.78% [kernel] [k] ixgbe_clean_rx_irq
2.74% [kernel] [k] fib_table_lookup
2.46% [kernel] [k] csum_partial_copy_generic
2.15% [kernel] [k] memcpy
1.66% [kernel] [k] ixgbe_xmit_frame_ring
1.55% [kernel] [k] __netif_receive_skb_core
1.42% [kernel] [k] __dev_queue_xmit
1.35% [kernel] [k] ip_route_input_noref
1.31% [kernel] [k] timerqueue_add
1.17% [kernel] [k] tcp_mt
1.17% [kernel] [k] ixgbe_poll
1.15% [kernel] [k] ip_finish_output
1.08% [kernel] [k] dev_gro_receive
1.06% [kernel] [k] nf_iterate
1.01% [kernel] [k] _raw_spin_lock_irqsave
1.01% [kernel] [k] check_leaf.isra.6
0.86% [kernel] [k] ip_rcv
0.82% [kernel] [k] cpu_startup_entry
0.80% [kernel] [k] dev_hard_start_xmit
0.80% [kernel] [k] kmem_cache_alloc
0.76% [kernel] [k] menu_select
0.74% [kernel] [k] u32_classify
0.73% [kernel] [k] put_page
0.69% [kernel] [k] ppp_xmit_process
0.68% [kernel] [k] irq_entries_start
0.67% [kernel] [k] tbf_dequeue
0.63% [kernel] [k] ip_forward
0.61% [kernel] [k] qdisc_dequeue_head
0.60% [kernel] [k] __local_bh_enable_ip
0.60% [kernel] [k] _raw_spin_lock_bh
0.59% [kernel] [k] __inet_lookup_established
0.59% [kernel] [k] apic_timer_interrupt
0.58% [kernel] [k] inet_gro_receive
0.55% [kernel] [k] __build_skb
0.51% [kernel] [k] rb_erase
0.50% [kernel] [k] kmem_cache_free
0.48% [kernel] [k] _raw_spin_unlock_irqrestore
0.45% [kernel] [k] int_sqrt
0.43% [kernel] [k] __netdev_pick_tx
0.42% [kernel] [k] nf_hook_slow
Re: Zombie process using 60% CPU
I think ht visibility fake load and not recommended use on soft router. Maybe spread network card interrupts to CPU1,CPU2,CPU3,CPU4,CPU5,CPU6,CPU7. CPU0 interrupts use only for system needs.
Re: Zombie process using 60% CPU
Hello!
How is it possible to improve the productive capacity of pppoe_rcv?
How is it possible to improve the productive capacity of pppoe_rcv?
Re: Zombie process using 60% CPU
Hi, what linux kernel version you are using? Maybe kernel update help you.
Re: Zombie process using 60% CPU
4.7.0-0.bpo.1-amd64
Re: Zombie process using 60% CPU
What about 4.8.last or 4.9? I'll try too. I see a lot of network fixes in last commits.
Maybe is acceptable since is balancing between the cores.
Another question. If I use ipoe will I got more performance? (considering arp cpu, vlans...)
Maybe is acceptable since is balancing between the cores.
Another question. If I use ipoe will I got more performance? (considering arp cpu, vlans...)