On 13.07.2011 00:26, Alexander Duyck wrote: > Looking over your lspci and dmesg log it does look like the PCIe resources > should be sufficient. The other possibility is that the adapter is being > denied memory resources by the CPU going into a package C-state and as such > disabling the memory controller. That is the reason. Both "idle=poll" and limiting to C1 max state in bios solve the problem. I believe this information must present in driver's readme. Thank you a lot.
> One way to verify this is to boot your kernel with the parameter idle=poll. > If this configuration does not drop any packets then the issue is likely due > to the system going into a package C-state. If this is the case you may want > to check your BIOS and see what options you have for reducing the sleep > states > supported on your system. > > Thanks, > > Alex > > On 07/12/2011 09:57 AM, Bokhan Artem wrote: >> Width x4 shoud be enough for 1Gbit/s. lspci and dmesg are attached. >> >> ps. The MB has x8 PCI-E 2.0 (supermicro X8SI6-F) >> >> 12.07.2011 22:54, Alexander Duyck пишет: >>> The rx_fifo_errors seem to be the result of the rx_missed_errors. >>> Looking over the statistics it would seem like you are seeing >>> rx_missed_errors but no rx_no_buffer_count errors. Typically this >>> combination will indicate a lack of sufficient PCIe bandwidth. >>> >>> Could you provide a "lspci -vvv" for the two ports, and perhaps even >>> the entire system. Also the output from dmesg would be useful as it >>> may provide additional information. >>> >>> I'm suspecting we may see that the adapter is on a x2 or x1 PCIe link. >>> If that is the case it will not be able to handle bursty traffic very >>> well and may drop packets when receiving a surge of traffic. >>> >>> Thanks, >>> >>> Alex >>> >>> On 07/12/2011 01:23 AM, Bokhan Artem wrote: >>>> Hello. >>>> >>>> I want to understand if it could be possible to avoid rx_fifo_errors? >>>> >>>> Network router, about 30 kpps on RX, 200 Mbit/s average on RX, about >>>> 20-30% cpu >>>> busy. Kernel: 2.6.32-32-server (ubuntu 10lts) >>>> >>>> modinfo igb >>>> filename: /lib/modules/2.6.32-32-server/kernel/igb/igb.ko >>>> version: 3.0.22 >>>> license: GPL >>>> description: Intel(R) Gigabit Ethernet Network Driver >>>> author: Intel Corporation,<[email protected]> >>>> srcversion: 45B8078075068728A5A5573 >>>> >>>> cat /etc/modprobe.d/igb.conf >>>> options igb RSS=4,4 QueuePairs=0,0 InterruptThrottleRate=100000,100000 >>>> >>>> >>>> 01:00.0 Ethernet controller: Intel Corporation 82576 Gigabit Network >>>> Connection >>>> (rev 01) >>>> Subsystem: Intel Corporation Device a03c >>>> Flags: bus master, fast devsel, latency 0, IRQ 16 >>>> Memory at fabe0000 (32-bit, non-prefetchable) [size=128K] >>>> Memory at fb000000 (32-bit, non-prefetchable) [size=4M] >>>> I/O ports at bc00 [size=32] >>>> Memory at fabdc000 (32-bit, non-prefetchable) [size=16K] >>>> Expansion ROM at fac00000 [disabled] [size=4M] >>>> Capabilities: [40] Power Management version 3 >>>> Capabilities: [50] Message Signalled Interrupts: Mask+ 64bit+ >>>> Queue=0/0 Enable- >>>> Capabilities: [70] MSI-X: Enable+ Mask- TabSize=10 >>>> Capabilities: [a0] Express Endpoint, MSI 00 >>>> Capabilities: [100] Advanced Error Reporting<?> >>>> Capabilities: [140] Device Serial Number 40-b9-8c-ff-ff-21-1b-00 >>>> Capabilities: [150] #0e >>>> Capabilities: [160] #10 >>>> Kernel driver in use: igb >>>> Kernel modules: igb >>>> >>>> 01:00.1 Ethernet controller: Intel Corporation 82576 Gigabit Network >>>> Connection >>>> (rev 01) >>>> Subsystem: Intel Corporation Device a03c >>>> Flags: bus master, fast devsel, latency 0, IRQ 17 >>>> Memory at f9fe0000 (32-bit, non-prefetchable) [size=128K] >>>> Memory at fa400000 (32-bit, non-prefetchable) [size=4M] >>>> I/O ports at b880 [size=32] >>>> Memory at f9fdc000 (32-bit, non-prefetchable) [size=16K] >>>> Expansion ROM at fa000000 [disabled] [size=4M] >>>> Capabilities: [40] Power Management version 3 >>>> Capabilities: [50] Message Signalled Interrupts: Mask+ 64bit+ >>>> Queue=0/0 Enable- >>>> Capabilities: [70] MSI-X: Enable+ Mask- TabSize=10 >>>> Capabilities: [a0] Express Endpoint, MSI 00 >>>> Capabilities: [100] Advanced Error Reporting<?> >>>> Capabilities: [140] Device Serial Number 40-b9-8c-ff-ff-21-1b-00 >>>> Capabilities: [150] #0e >>>> Capabilities: [160] #10 >>>> Kernel driver in use: igb >>>> Kernel modules: igb >>>> >>>> ethtool -S eth1 >>>> NIC statistics: >>>> rx_packets: 188369369 >>>> tx_packets: 147990215 >>>> rx_bytes: 201929045712 >>>> tx_bytes: 33522657344 >>>> rx_broadcast: 887 >>>> tx_broadcast: 2542 >>>> rx_multicast: 21892 >>>> tx_multicast: 7705 >>>> multicast: 21892 >>>> collisions: 0 >>>> rx_crc_errors: 0 >>>> rx_no_buffer_count: 0 >>>> rx_missed_errors: 602 >>>> tx_aborted_errors: 0 >>>> tx_carrier_errors: 0 >>>> tx_window_errors: 0 >>>> tx_abort_late_coll: 0 >>>> tx_deferred_ok: 0 >>>> tx_single_coll_ok: 0 >>>> tx_multi_coll_ok: 0 >>>> tx_timeout_count: 0 >>>> rx_long_length_errors: 0 >>>> rx_short_length_errors: 0 >>>> rx_align_errors: 0 >>>> tx_tcp_seg_good: 651453 >>>> tx_tcp_seg_failed: 0 >>>> rx_flow_control_xon: 0 >>>> rx_flow_control_xoff: 0 >>>> tx_flow_control_xon: 0 >>>> tx_flow_control_xoff: 0 >>>> rx_long_byte_count: 201929045712 >>>> tx_dma_out_of_sync: 0 >>>> tx_smbus: 0 >>>> rx_smbus: 0 >>>> dropped_smbus: 0 >>>> os2bmc_rx_by_bmc: 0 >>>> os2bmc_tx_by_bmc: 0 >>>> os2bmc_tx_by_host: 0 >>>> os2bmc_rx_by_host: 0 >>>> rx_errors: 0 >>>> tx_errors: 0 >>>> tx_dropped: 0 >>>> rx_length_errors: 0 >>>> rx_over_errors: 0 >>>> rx_frame_errors: 0 >>>> rx_fifo_errors: 602 >>>> tx_fifo_errors: 0 >>>> tx_heartbeat_errors: 0 >>>> tx_queue_0_packets: 36702209 >>>> tx_queue_0_bytes: 8074726795 >>>> tx_queue_0_restart: 0 >>>> tx_queue_1_packets: 38005442 >>>> tx_queue_1_bytes: 8377818990 >>>> tx_queue_1_restart: 0 >>>> tx_queue_2_packets: 36434405 >>>> tx_queue_2_bytes: 8067577883 >>>> tx_queue_2_restart: 0 >>>> tx_queue_3_packets: 36848159 >>>> tx_queue_3_bytes: 7977808543 >>>> tx_queue_3_restart: 0 >>>> rx_queue_0_packets: 47764596 >>>> rx_queue_0_bytes: 51072172246 >>>> rx_queue_0_drops: 0 >>>> rx_queue_0_csum_err: 0 >>>> rx_queue_0_alloc_failed: 0 >>>> rx_queue_1_packets: 49050786 >>>> rx_queue_1_bytes: 53273754217 >>>> rx_queue_1_drops: 0 >>>> rx_queue_1_csum_err: 0 >>>> rx_queue_1_alloc_failed: 0 >>>> rx_queue_2_packets: 46223688 >>>> rx_queue_2_bytes: 48353169643 >>>> rx_queue_2_drops: 0 >>>> rx_queue_2_csum_err: 0 >>>> rx_queue_2_alloc_failed: 0 >>>> rx_queue_3_packets: 45330301 >>>> rx_queue_3_bytes: 48476472794 >>>> rx_queue_3_drops: 0 >>>> rx_queue_3_csum_err: 0 >>>> rx_queue_3_alloc_failed: 0 >>>> >>>> ethtool -k eth1 >>>> Offload parameters for eth1: >>>> rx-checksumming: off >>>> tx-checksumming: on >>>> scatter-gather: on >>>> tcp-segmentation-offload: on >>>> udp-fragmentation-offload: off >>>> generic-segmentation-offload: on >>>> generic-receive-offload: off >>>> large-receive-offload: off >>>> >>>> ethtool -g eth1 >>>> Ring parameters for eth1: >>>> Pre-set maximums: >>>> RX: 4096 >>>> RX Mini: 0 >>>> RX Jumbo: 0 >>>> TX: 4096 >>>> Current hardware settings: >>>> RX: 4096 >>>> RX Mini: 0 >>>> RX Jumbo: 0 >>>> TX: 4096 >>>> >>>> >>>> ------------------------------------------------------------------------------ >>>> >>>> All of the data generated in your IT infrastructure is seriously >>>> valuable. >>>> Why? It contains a definitive record of application performance, >>>> security >>>> threats, fraudulent activity, and more. Splunk takes this data and makes >>>> sense of it. IT sense. And common sense. >>>> http://p.sf.net/sfu/splunk-d2d-c2 >>>> _______________________________________________ >>>> E1000-devel mailing list >>>> [email protected] >>>> https://lists.sourceforge.net/lists/listinfo/e1000-devel >>>> To learn more about Intel® Ethernet, visit >>>> http://communities.intel.com/community/wired >>> >> >> > ------------------------------------------------------------------------------ AppSumo Presents a FREE Video for the SourceForge Community by Eric Ries, the creator of the Lean Startup Methodology on "Lean Startup Secrets Revealed." This video shows you how to validate your ideas, optimize your ideas and identify your business strategy. http://p.sf.net/sfu/appsumosfdev2dev _______________________________________________ E1000-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/e1000-devel To learn more about Intel® Ethernet, visit http://communities.intel.com/community/wired
