[c-nsp] 2811 router doesn't recognize T-1 WICs?

Dave Temkin dave at ordinaryworld.com
Wed Jan 26 11:46:23 EST 2005


On Wed, 26 Jan 2005, Rodney Dunn wrote:

> On Wed, Jan 26, 2005 at 10:46:29AM -0500, Dave Temkin wrote:
> > FWIW, I've had nothing but great success with the 2811's/2821's (and
> > 3845's) thus far.  I've got a couple of the WIC-1DSU's, but mostly have
> > VWIC-2MFT-T1's and they work flawlessly thus far.  I have about 20 T1's in
> > production using them.
> >
> > They can be had on eBay for $170, so you might want to consider to
> > swapping to those instead of switching platforms.  Seems like the -V2's
> > have more production problems than they're worth because they're new.
>
> Buying on eBay doesn't pay my salary and if I don't get any
> money I can't work here and if I don't work here I can't
> help you as much. :)   /*a little humor*/

True, but that's the trade-off you guys take for selling new routers
(2800's/3800's/etc.) :-)  Plus, my ridiculous amount of maintenance that I
pay a year is what technically pays *your* salary, right? :-)


>
> >
> > Only complaint I have thus far on the 28xx is that when you do a write mem
> > it can cause ISIS adjacencies to flap if the hello-interval is set to
> > minimal.  Still need to get a ddts opened on that.
>
> Can you tell me how low you set the timers?

I used isis hello-interval minimal, so 1 second

>
> Try tweaking this down by increments of 20
> and see if it helps:
>
> 201_(config)#process-max-time ?
>   <20-200>  The number of millisecs before voluntary suspend
>
>
> Let me know if that helps you.
>
> Basically a process can run for 200 msec before doing
> a voluntary suspend.  If your timers are too low you
> can't process them fast enough.
>

So right now if the write runs for more than, say, 100msec and the hello
is coming up in the next 10msec, it won't run the hello until the write
hits 200msec?

Thanks,
-Dave



> The answer to this will be BFD.
>
> Rodney
>
>
> >
> > -Dave
> >
> > On Wed, 26 Jan 2005, John Neiberger wrote:
> >
> > > Dave,
> > >
> > > That's a great idea. If you ask them to make it viewable then I'll be
> > > able to track the progress of these bugs. I'd like to use the 2811s and
> > > those WICs but now I'm wondering if I should cancel my order and stick
> > > with a 2620XM instead. Does anyone know if the V2 WICs are experiencing
> > > problems when used in the 2600XM series or does this just happen in the
> > > 2800 series routers?
> > >
> > > Thanks,
> > > John
> > > --
> > >
> > > >>> Rodney Dunn <rodunn at cisco.com> 1/26/05 8:25:22 AM >>>
> > > For the record.
> > >
> > > If you have a case open and the bug is not
> > > viewable on CCO then ask the TAC engineer
> > > to make it viewable.
> > >
> > > There are usually only two reasons a bug
> > > is not viewable on CCO.
> > >
> > > a) PSIRT
> > > b) It was found internally in testing and was thought
> > >    to only exist in images that were never released to customers.
> > >    If it's determined it was in an image released to a customer
> > >    the correct procedure is to modify that bug so it is visible
> > >    on CCO with the correct enclosures (ie: Release-note).
> > >
> > > ...or there is a 3rd reason and someone just messed up.
> > >
> > > Bottom line, if you encounter a bug not viewable on
> > > CCO that you are hitting ask them to make it viewable.
> > >
> > > Rodney
> > >
> > >
> > > On Wed, Jan 26, 2005 at 09:14:48AM -0600, MADMAN wrote:
> > > >
> > > >
> > > >    Concerning the WIC-1DSU-T1-V2 we have been ecperiencing many
> > > problems
> > > > and have a TAC case opened.  So far we have identifed twi issues that
> > >
> > > > Cisco has bugs id's assigned to.
> > > >
> > > > - WIC runs for 2 hours, 17 minutes then goes down
> > > > - High number of T1 errors seen with V2 card, no errors seen on same
> > >
> > > > line with V1 card
> > > >
> > > >    These are Cisco internal bugs so I can't provide a bug ID.  Though
> > >
> > > > these specific cases may not reflect your experiences me thinks the
> > > V2
> > > > has multiple issues.
> > > >
> > > >    Dave
> > > >
> > > > Chris Cappuccio wrote:
> > > >
> > > > > Open a case with the Cisco TAC.  Since your 2811 is probably pretty
> > > new,
> > > > > you shouldn't even need a service contract (yet).  They'll send you
> > > replacement
> > > > > WICs and test to figure out why your current ones failed.
> > > > >
> > > > > So far, with the new 2811s, I have seen an instance where one
> > > customer bought
> > > > > a mix of six 2811 and 2821 routers, and six WIC-1DSU-T1-V2s, only
> > > to find
> > > > > that three out of the six V2 WICs DID NOT WORK.  The broken three
> > > ONLY worked
> > > > > in slot 0, and often times the T1s did not work properly even in
> > > that instance
> > > > > (it would go down and a reload was the only way to bring it back
> > > up).
> > > > >
> > > > > The good three WICs worked fine, in any slot.  The kicker is that
> > > the six
> > > > > WICs had _SEQUENTIAL_ serial numbers, so for half of them to be
> > > "bad" in the
> > > > > 2811s was definitely not expected.  They seemed to work fine in
> > > 3600 and
> > > > > other older routers.
> > > > >
> > > > > Same customer, same routers, they also had a problem with a 2811
> > > that
> > > > > is even more worrisome.  One 2811, on a particular T1, would not
> > > stay
> > > > > up, "Line protocol is down", unless you set hdlc keepalive to 13 or
> > > something
> > > > > else odd.  Plug in a 2500 with external CSU or 2600 with a WIC and
> > > the T1
> > > > > works just fine with completely default settings.  Ironically,
> > > other
> > > > > 2811s work fine at this location with default settings.  All 2811s
> > > had
> > > > > the same IOS images (not that there is much choice for these new
> > > boxes)
> > > > >
> > > > > Yep, there are strange issues with these new boxes.  Given Cisco's
> > > marketing
> > > > > push, I would imagine most early adopters are using them primarily
> > > in
> > > > > voice applications and not data T1s.  Perhaps they were better
> > > debugged there.
> > > > >
> > > > > All I have left is to say "Way to go Cisco!!" for the broken T1
> > > technology -
> > > > > I know it's highly complicated, delicate, top precision work, and
> > > that the
> > > > > utmost time and care goes into testing such obvious
> > > configurations...err..uhh..
> > > > > wait a minute, what was I saying ???
> > > > >
> > > > > Jay Hennigan [jay at west.net] wrote:
> > > > >
> > > > >>Hello fellow Cisco-NSP folk.
> > > > >>
> > > > >>I have a brand new 2811 router which according to the Cisco DPRG
> > > > >>configurator is supposed to take the WIC-1DSU-T1-V2 modular WICs
> > > > >>as opposed to the non-V2 type.
> > > > >>
> > > > >>So we paid the higher price for brand-spanking-new V2 modules, and
> > > they
> > > > >>don't show up.  "sho diag" gives:
> > > > >>
> > > > >>        WIC Slot 0:
> > > > >>        Unknown WAN daughter card
> > > > >>        WIC module not supported/disabled in this slot
> > > > >>        Hardware revision 64.0          Board revision >1
> > > > >>        Serial number     1379991808    Part number
> > > 800-7441222-79
> > > > >>        Test history      0x43          RMA number     48-55-51
> > > > >>        Connector type    PCI
> > > > >>        EEPROM format version 1
> > > > >>        EEPROM contents (hex):
> > > > >>          0x20: 01 80 40 00 52 41 01 00 C1 8B 46 4F 43 30 37 33
> > > > >>          0x30: 31 31 41 36 4A 82 49 20 9A 03 42 41 30 03 00 81
> > > > >>
> > > > >>Tried two different WICs in all slots with similar results.
> > > > >>
> > > > >>Is there some magic needed to get these recognized or do I have a
> > > hardware
> > > > >>issue?  The box is running IP-Base 12.3(11)T2, same results with
> > > 12.3(8)T.
> > > > >>
> > > > >>--
> > > > >>Jay Hennigan - CCIE #7880 - Network Administration - jay at west.net
> > > > >>WestNet:  Connecting you to the planet.  805 884-6323      WB6RDV
> > > > >>NetLojix Communications, Inc.  -  http://www.netlojix.com/
> > > > >>_______________________________________________
> > > > >>cisco-nsp mailing list  cisco-nsp at puck.nether.net
> > > > >>https://puck.nether.net/mailman/listinfo/cisco-nsp
> > > > >>archive at http://puck.nether.net/pipermail/cisco-nsp/
> > > > >
> > > > >
> > > >
> > > > --
> > > > David Madland
> > > > CCIE# 2016
> > > > Sr. Network Engineer
> > > > Qwest Communications
> > > > 612-664-3367
> > > >
> > > > "Emotion should reflect reason not guide it"
> > > > _______________________________________________
> > > > cisco-nsp mailing list  cisco-nsp at puck.nether.net
> > > > https://puck.nether.net/mailman/listinfo/cisco-nsp
> > > > archive at http://puck.nether.net/pipermail/cisco-nsp/
> > > _______________________________________________
> > > cisco-nsp mailing list  cisco-nsp at puck.nether.net
> > > https://puck.nether.net/mailman/listinfo/cisco-nsp
> > > archive at http://puck.nether.net/pipermail/cisco-nsp/
> > > _______________________________________________
> > > cisco-nsp mailing list  cisco-nsp at puck.nether.net
> > > https://puck.nether.net/mailman/listinfo/cisco-nsp
> > > archive at http://puck.nether.net/pipermail/cisco-nsp/
> > >
>


More information about the cisco-nsp mailing list