[f-nsp] High CPU on LP

[Michael Gehrmann]

Do you use this feature?

ip tcp [ adjust-mss max-segment-size ]

On 30 March 2017 at 16:18, Daniel Zilla <dzilla833 at yandex.com> wrote:

> Hello List
>
> LP CPU is stuck at 96-100% when receiving traffic over a configured GRE
> tunnel
>
> MLX-4
> NI-MLX-10Gx4
>
> Debug packet capture shows many of these
>
> [ppcr_tx_packet] ACTION: Drop packet(reason: Layer 3 invalid FID(PFE))
> [ppcr_rx_packet]: Packet received
> Time stamp : 188 day(s) 12h 45m 15s:,
> TM Header: [ 8016 1000 0000 ]
> Type: Fabric Unicast(0x00000008) Size: 88 Parity: 1 Src IF: 0
> Src Fap: 0 Dest Port: 0  Src Type: 0 Class: 0x00000000
> **********************************************************************
> Packet size: 82, XPP reason code: 0x00006868
> FID     = 0x0001
>   Offset  = 0x10
>   VLAN    = 100(0x0064)
>   CAM     = 0x135ad(R)
>   SFLOW   = 0
>   DBL TAG = 0
>
> Pri CPU MON SRC   PType US BRD DAV SAV DPV SV ER TXA SAS Tag MVID
> 0   1   0   1/1   0     1  0   0   0   1   0  0  1   1   1   0
>
> The issue only occurs when traffic is sent over the GRE tunnel otherwise
> CPU is 1 - 2 %
>
> Management CPU is low all the time, LP is the only problematic
>
> Show tasks indicates LP-RX-Pak is 96% : main is 2%
>
> No route-only mode is enabled on this chassis.
>
> Cheers
>
> _______________________________________________
> foundry-nsp mailing list
> foundry-nsp at puck.nether.net
> http://puck.nether.net/mailman/listinfo/foundry-nsp
>



-- 
Michael Gehrmann
Senior Network Engineer - Atlassian
m: +61 407 570 658
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://puck.nether.net/pipermail/foundry-nsp/attachments/20170330/5e432f24/attachment.html>