[f-nsp] Measurement of BigIron virtual interfaces

Brent Van Dussen vandusb at attens.com
Tue Oct 26 15:17:36 EDT 2004


A 'show modules' will give you the info:

Here's a bi8k with M4 modules:

     Module                                 Status    Ports Starting MAC
S1: B4GMR4 M4 Management Module, SYSIF 2   M4, ACTIV   4   0004.8048.6e00
S2: B4GMR4 M4 Management Module, SYSIF 2   M4, STAND   4   0004.8048.6e20
S3: B8G Fiber Switch Module, SYSIF 2       OK          8   0004.8048.6e40
S4: B8G Fiber Switch Module, SYSIF 2       OK          8   0004.8048.6e60

Here's a bi8k with Jetcore Modules:

     Module                                 Status    Ports Starting MAC
S1: J-BxGMR4 JetCore Management Module, SY M4, ACTIV   8   000c.db48.9300
S2: J-BxGMR4 JetCore Management Module, SY M4, STAND   8   000c.db48.9320
S3: J-BxG JetCore Gig Fiber Module, SYSIF  OK          8   000c.db48.9340
S4: J-BxG JetCore Gig Fiber Module, SYSIF  OK          8   000c.db48.9300
S5: J-BxG JetCore Gig Fiber Module, SYSIF  OK          8   000c.db48.9300
S6:
S7: J-B48E-A JetCore Copper E Module, SYSI OK         48   000c.db48.9300
S8:

-Brent


At 11:44 AM 10/26/2004, Sam Stickland wrote:
>Hmm, then it seems we have been, err. missupplied somewhere along the 
>line. This card was listed as a J-BxGMR4 when the provisioning request was 
>originally made, which ought to make it Jetcore based but evidently this 
>isn't one of them.
>
>What would that M4 show up as under a sh ver?
>
>We
>
>Sam
>
>On Tue, 26 Oct 2004, Jonas Frey wrote:
>
>>Sam,
>>
>>you have a old M4 based on the ironcore chipset (see my last mail to the
>>list), this doesnt support sflow. You can configure it but it will NOT
>>work.
>>
>>A VM1 *does* support sflow.
>>
>>Regards,
>>Jonas
>>
>>
>>On Tue, 2004-10-26 at 18:13, Sam Stickland wrote:
>>>On Mon, 25 Oct 2004, Ray Taft wrote:
>>>
>>>>No, you can not monitor a VE interface. It will report back zero usage.
>>>>Bummer, I know.
>>>>
>>>>Ray
>>>
>>>That is indeed a pain.
>>>
>>>Does anyone know if the lack of VE interface counters is a hardware
>>>problem? Does it change with a VM-1 card installed?
>>>
>>>I've started looking at how to extract the information from sFlow, but
>>>annoyingly I can't even seem to get sFlow exporting working.
>>>
>>>SSH at foundry1#sh run | inc sflow
>>>sflow enable
>>>sflow sample 512
>>>sflow polling-interval 30
>>>sflow destination w.x.y.z
>>>
>>>SSH at foundry1#conf t
>>>SSH at foundry1(config)#int e1/3
>>>SSH at foundry1(config-if-e1000-1/3)#sflow forwarding
>>>SSH at foundry1(config-if-e1000-1/3)#exit
>>>SSH at foundry1(config)#exit
>>>
>>>SSH at foundry1#sh sflow
>>>sFlow services are enabled.
>>>sFlow agent IP address: 10.134.221.4
>>>Collector IP w.x.y.z, UDP 6343
>>>Polling interval is 30 seconds.
>>>Configured default sampling rate: 1 per 512 packets.
>>>Actual default sampling rate: 1 per 512 packets.
>>>0 UDP packets exported
>>>0 sFlow samples collected.
>>>No ports configured for sFlow sampling.
>>>Module Sampling Rates
>>>
>>>SSH at foundry1#sh run int e1/3
>>>!Building configuration...
>>>!Current configuration : 47 bytes
>>>interface ethernet 1/3
>>>   port-name Gi0/1 Net1
>>>!
>>>
>>>So basically it's just ignoring the sFlow port configuration. I've read
>>>through the Foundry docs and can't seem to figure out why. The interface
>>>e1/3 is one of the GE ports on the B4GMR4 card, and the box is running
>>>07.7.01T53 (B2R07701).
>>>
>>>Feels like I'm missing something very obvious, but I can't see what.
>>>
>>>Sam
>>>
>>>>-----Original Message-----
>>>>From: foundry-nsp-bounces at puck.nether.net
>>>>[mailto:foundry-nsp-bounces at puck.nether.net] On Behalf Of Jonas Frey
>>>>Sent: Monday, October 25, 2004 12:07 PM
>>>>To: Cliff Fogle
>>>>Cc: foundry-nsp at puck.nether.net
>>>>Subject: RE: [f-nsp] Measurement of BigIron virtual interfaces
>>>>
>>>>Hi,
>>>>
>>>>please bare in mind: sflow is only supported on the jetiron m4 boards,
>>>>not on the (old) ironcore ones.
>>>>I dont know about ve monitoring/accounting.
>>>>
>>>>Btw, i got some foundry equipment to sell...if anyone is interessted:
>>>>B2P622-SM-IR
>>>>M4R4G (somehow defective, tends to crash when running bgp/using alot of
>>>>routes)
>>>>
>>>>
>>>>Regards,
>>>>Jonas
>>>>
>>>>On Mon, 2004-10-25 at 18:57, Cliff Fogle wrote:
>>>>>NTOP kind of has sflow support.  It really only gives 'top talkers'
>>>>>info.  The best thing out there at this time is really InMon's traffic
>>>>>server:
>>>>>
>>>>>http://www.inmon.com/products/trafficserver.php
>>>>>
>>>>>(warning:  There is a free demo avaialble and it's addictive)
>>>>>
>>>>>Foundry is apparently building much better sflow support into INM 1.8,
>>>>>but I have yet to see anything other than a PowerPoint mock up of it.
>>>>>
>>>>>-----Original Message-----
>>>>>From: foundry-nsp-bounces at puck.nether.net
>>>>>[mailto:foundry-nsp-bounces at puck.nether.net] On Behalf Of Sam Stickland
>>>>>Sent: Monday, October 25, 2004 6:47 AM
>>>>>To: foundry-nsp at puck.nether.net
>>>>>Subject: [f-nsp] Measurement of BigIron virtual interfaces
>>>>>
>>>>>Hi,
>>>>>
>>>>>Is it possible to measure the virtual interface bandwidth stats on the
>>>>>BigIrons (Mmgt4)? I notice that the interface OID's appear to be
>>>>>readable but the counters never seem to increment (running 07.7.01T53)
>>>>>
>>>>>Can this information be read via SNMP. If not, a cursorsy glance at the
>>>>>sFlow specs seems to suggest is it possible to measure this information
>>>>>using sFlow. If so are there any opensource or reasonably lightweight
>>>>>tools that can process the information in this regard?
>>>>>
>>>>>Thanks,
>>>>>
>>>>>Sam
>>>>>_______________________________________________
>>>>>foundry-nsp mailing list
>>>>>foundry-nsp at puck.nether.net
>>>>>http://puck.nether.net/mailman/listinfo/foundry-nsp
>>>>>
>>>>>
>>>>>
>>>>>_______________________________________________
>>>>>foundry-nsp mailing list
>>>>>foundry-nsp at puck.nether.net
>>>>>http://puck.nether.net/mailman/listinfo/foundry-nsp
>>>>
>>>>_______________________________________________
>>>>foundry-nsp mailing list
>>>>foundry-nsp at puck.nether.net
>>>>http://puck.nether.net/mailman/listinfo/foundry-nsp
>>>>
>>>>_______________________________________________
>>>>foundry-nsp mailing list
>>>>foundry-nsp at puck.nether.net
>>>>http://puck.nether.net/mailman/listinfo/foundry-nsp
>>
>>
>_______________________________________________
>foundry-nsp mailing list
>foundry-nsp at puck.nether.net
>http://puck.nether.net/mailman/listinfo/foundry-nsp





More information about the foundry-nsp mailing list