[afnog] L2 channel errors
ibtisam jamal
ibty.jamal at gmail.com
Fri Nov 4 08:27:32 UTC 2011
Hi ,
for this Ge PIC oversubscription 4:1 the utilization is actualy around
920Mbps ,and i can see the RED drops ,does this mean that the 1 Ge is
actualy working at around 800Mbps line rate ?
Regards,
Ibtisam
On Wed, Nov 2, 2011 at 10:32 PM, ibtisam jamal <ibty.jamal at gmail.com> wrote:
> Hi,
>
> Please see my response below...
>
> On Wed, Nov 2, 2011 at 3:32 PM, Mark Tinka <mtinka at globaltransit.net>wrote:
>
>> On Wednesday, November 02, 2011 05:26:53 PM ibtisam jamal
>> wrote:
>>
>> > from 4 ge interfaces on one ge PIC...
>>
>> Sounds like you're talking about the 4-port Gig-E PIC. Just
>> not sure what Juniper platform you're talking about.>> Juniper m10i
>>
>>
>> > all 4 interfaces show
>> > the errors ,while only 2 of them are configured as
>> > trunks the other 2 have only one vlan configured
>> > each.All the 4 interfaces have the L2 errors.Last night
>> > we tried to swap the PIC but today the same errors.
>>
>> The 4-port Gig-E PIC, both for the M7i/M10i as well as the
>> larger models, are oversubscribed 4:1. This means that
>> effectively, with all ports running at line rate, you can
>> only achieve 25% of 1Gbps on each port, as there is only a
>> 1Gbps connection to the PFE for that PIC.>>>is there a way to check this
>> 1Gbps connection utilization .
>>
> Would you know what these checksum errors mean and the discards.
>
> @@@@@@@ show pfe statistics traffic
> Packet Forwarding Engine traffic statistics:
> Input packets: 2826536051484 398994 pps
> Output packets: 2819386113592 398234 pps
> Packet Forwarding Engine local traffic statistics:
> Local packets input : 3498362275
> Local packets output : 1091783398
> Software input control plane drops : 0
> Software input high drops : 26
> Software input medium drops : 18346
> Software input low drops : 0
> Software output drops : 14650
> Hardware input drops : 7816
> Packet Forwarding Engine local protocol statistics:
> HDLC keepalives : 464379
> ATM OAM : 0
> Frame Relay LMI : 0
> PPP LCP/NCP : 1059990
> OSPF hello : 20171430
> OSPF3 hello : 0
> RSVP hello : 8061634
> LDP hello : 0
> BFD : 206871486
> IS-IS IIH : 0
> LACP : 0
> ARP : 2151007242
> ETHER OAM : 0
> Unknown : 162571
> Packet Forwarding Engine hardware discard statistics:
> Timeout : 0
> Truncated key : 2
> Bits to test : 0
> Data error : 0
> Stack underflow : 0
> Stack overflow : 0
> Normal discard : 4589614365
> Extended discard : 5763
> Invalid interface : 0
> Info cell drops : 0
> Fabric drops : 0
> Packet Forwarding Engine Input IPv4 Header Checksum Error and Output MTU
> Error statistics:
> Input Checksum : 8610737561
> Output MTU : 5427563
>
>
>
>>
>> The reason I bring that up is I'm not quite familiar with
>> this PIC's architecture - I don't know whether the errors
>> you're seeing are reported for all ports, or for each port.>>>>>errors on
>> all ports on this PIC
>>
>>
>> You say that 2x of the ports are running 802.1Q. Please
>> confirm whether all VLAN's coming into the router from the
>> switch are present on both sides of the link.>>>>>>will confirm when i
>> get connection and get back to you.
>>
>>
>> Also, you mention that the other two remaining ports have
>> only one VLAN each, but no 802.1Q enabled. I don't see how
>> that's possible, as configurations of VLAN's on the PIC's
>> requires 802.1Q ('vlan-tagging' in Junos speak) to be
>> enabled. >>>yes vlan-tagging is enabled.
>>
>>
>> > We also see packet drops under the class-of-service
>>
>> I think let's look at that separately, in your e-mail on the
>> same, so we don't get the thread confused :-).
>>
>> Mark.
>>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://afnog.org/pipermail/afnog/attachments/20111104/52dad08a/attachment.html>
More information about the afnog
mailing list