- Test - lists.nanog.org

Re: weird BGP cisco-ism? [problem resolved]
by Danny McPherson 11 Jul '97

11 Jul '97

> > Try it w/o the summary-only stuff. > > Didn't help at all. > > > > > I don't use summary-only, or aggregate-address, or network XXX at > > all. I redis my connected & static routes into bgp and arrange > > for my global routes to make it beyond my AS. I also use pull up > > static routes to null0 for all of my blocks. > > I redistribute in a similar way, and have the statics to null0 as well. > > Still insisted on flapping at least twice every time it got the brief > hit on the /24 subnet coming from the outside world. > > At this point what needs to happen is for this to be recreated in > a lab somewhere and some config changes tried... I can't blow away > my customers one more time just to play with it. > > Some day we'll have routers that can tell you what they're doing, right? Because you're redistributing the route into BGP, when the "true" route becomes unavailable and the "nailed up" route kicks in, the route will flap. As Ravi mentioned: [snip] To avoid readvertising the same prefix, you might want to check if there is any other dynamic protocol that is installing that prefix in the routing table and removing it (could be a OSPF subnet...) ------- You could do a couple of things. 1.) Stop redistributing the more specifics into BGP and allow the "nailed up" routes to source the "aggregate" BGP advertisements. Something like I mentioned previously should suffice. Besides, there's really no need to inject the more specifics into BGP if you don't plan to advertise them to external peers .. per they're probably already being redistributed into your IGP. 2.) Configure a Loopback interface addressed out of the /20 network on the router which the summary-only statement exists so that a more specific of the network will always be available. I prefer sourcing advertisments via "Null 0" and "network" statments and don't like redistrbuting anything into BGP. -danny

1 0

Re: weird BGP cisco-ism? [problem resolved]
by matthew＠scruz.net 11 Jul '97

11 Jul '97

Original message <199707112126.RAA11467(a)home.partan.com> From: Andrew Partan <asp(a)partan.com> Date: Jul 11, 17:26 Subject: Re: weird BGP cisco-ism? [problem resolved] > > > provider, and we have "summary-only" set in our BGP configuration. Every > > Try it w/o the summary-only stuff. Didn't help at all. > > I don't use summary-only, or aggregate-address, or network XXX at > all. I redis my connected & static routes into bgp and arrange > for my global routes to make it beyond my AS. I also use pull up > static routes to null0 for all of my blocks. I redistribute in a similar way, and have the statics to null0 as well. Still insisted on flapping at least twice every time it got the brief hit on the /24 subnet coming from the outside world. At this point what needs to happen is for this to be recreated in a lab somewhere and some config changes tried... I can't blow away my customers one more time just to play with it. Some day we'll have routers that can tell you what they're doing, right? -matthew

3 2

Re: Mae-WEST
by Philip Steffora 11 Jul '97

11 Jul '97

I have a cage in the MFS Co-Lo and insisted on having a separate UPS from the main. While I never lost power today, my FDDI did go down. This gets me thinking about the backup generator that is supposed to be online. Why didn't the generator kick-in before the UPS went down? Perhaps the better question is if they have a functioning generator ... Anyone from MFS listening? -ps At 09:15 AM 7/11/97 -0700, you wrote: > > I would guess that power to the colo cages is down. > >Mike.

4 3

Re: Mae-WEST
by Barry A. Dykes 11 Jul '97

11 Jul '97

If I remember right, the 7000 series routers have the env sensors on the arbiter. I don't remember it being on a particular card, so it's not a simple swap :( Barry > At 03:40 PM 7/11/97 -0400, Golan Ben-Oni wrote: > >On Fri, 11 Jul 1997, Eric Kozowski wrote: > >Ironically, our temperature device is now broken in our Cisco. Has anyone > >else had similar problems with them? We now have a number of 7000 series > >routers which report off-the-wall readings. > > > >-Golan > > > > > > > > MFS is in the proccess of installing new cooling units on the roof. They > take up aprox. 1000sqft. So they are at least working on it. > > You should have seen the air lift :) > > > James McKenzie > mcs(a)innetix.com > Innetix, Inc. > (408) 993-9477 > http://www.innetix.com

1 0

Re: Mae-WEST
by Philip Steffora 11 Jul '97

11 Jul '97

Obviously, as a MFS customer that doesn't make me feel a whole bunch better about the situation in San Jose. -phil ================================== Philip Steffora steffora(a)pa.dec.com Operations Mgr. AltaVista - Digital Equip Corp ================================== At 11:25 AM 7/11/97 -0700, Steve Feldman wrote: >Yes, there's a generator. Unfortunately, it doesn't kick in when someone >trips or turns off the main breaker. (This is hearsay; I'm actually out >of the office and am getting my info second-hand.) > Steve

1 0

Re: Mae-WEST
by Sanjay Dani 11 Jul '97

11 Jul '97

>> This gets me thinking about the backup generator that is supposed to be >> online. >> >> Why didn't the generator kick-in before the UPS went down? >Or, lets say, an automatic network monitor (ALA SNMP) that tells them >when UPS power is about to run down (before the lights go out) so they >can roll their generator into place. It is amazing how many connectivity failures are power related. For an APC UPS, all it takes is a dedicated $500 sparcstation, the $99 powerchute UPS monitoring software, and a $99 modem attached to the sparc for sending out a text page every time the UPS coughs. Sorry for stating the obvious.

1 0

RE: Mae-WEST
by Steve Carter 11 Jul '97

11 Jul '97

The message I got was that a contractor tripped a breaker, the UPS was run down and out of juice and the generator took a while to kick in. The power outage took out Gigaswitch #1, effectivly dropping the whole MAE. When power was restored the switch had a memory problem that wouldn't allow the config to load. Memory fixed, config loaded, but now there is still a problem with a switch port (#4, MFS said) that has kept six of our peers down. A couple of suggestions/observations: o Monitor the UPS so it doesn't run out of juice. o Reconfigure the Gigaswitch topology so that one loss doesn't take out all peers. o MFS call handling at the customer service center was good, considering the number of calls they must have got. Shirley even took the time to call me directly with an update. Cool. Maybe MFS should consider a mail server that would send out periodic status updates to subscribers? ---------------------------------- > I have a cage in the MFS Co-Lo and insisted on having a separate UPS > from > the main. > While I never lost power today, my FDDI did go down. > > This gets me thinking about the backup generator that is supposed to > be > online. > > Why didn't the generator kick-in before the UPS went down? > > Perhaps the better question is if they have a functioning generator > ... > Anyone from MFS listening? > > -ps > > At 09:15 AM 7/11/97 -0700, you wrote: > > > > I would guess that power to the colo cages is down. > > > >Mike. >

1 0

RE: ATM vs. DS3
by Murray, Steve 11 Jul '97

11 Jul '97

Peter Kline wrote >Another way of looking at this is that we know an ATM PVC over a DS3 using >PLCP is configurable for a maximum of 96000 cells per second. 96000 cells >per second * 48 payload octets per cell * 8 bits per octet = 36864000 bps, >or 36.864 Mbps, not too far from what I observed above. With PLCP turned >off, the maximum PVC config was ~105000 cells per second, yielding ~40.3 Mpbs. The SAR function will add eight bytes to the payload. The eight bytes contain the CPCS-UU(1 bytes), CPI (1 byte), length of payload (two bytes) and a CRC (4 bytes). This additional overhead needs to be taken into account if you are segmenting small frames. The last cell after segmentation will have these eight bytes at the end of the cell plus any padding needed to fill out the cell. If you are segmenting a 96 byte payload for example you would think this would actually fit into two cells (using the 48 byte payload), but it really take three cells. The last cell would only contain padding and this overhead. Steve Murray Worldcom

1 0

MAE West is up again (fwd)
by Taner Halicioglu 11 Jul '97

11 Jul '97

Interesting that GS-1 isn't on the "main DC plant"... I thought this GS was an important one, considering all the OC-3's go to it :-) -Taner -- D. Taner Halicioglu taner(a)isi.net Programmer/Engineer/Sysadmin ISI / GlobalCenter Voice: +1 408 543 0313 Fax: +1 408 541 9878 PGP Fingerprint: 65 0D 03 A8 26 21 6D B8 23 3A D6 67 23 6E C0 36 ---------- Forwarded message ---------- Date: Fri, 11 Jul 1997 10:06:44 -0700 (PDT) From: Steve Feldman <feldman(a)mfst.com> To: mae-west(a)nsipo.arc.nasa.gov Subject: MAE West is up again MAE West should be alive now. If you're still seeing problems, please contact our NOC and let them know! Here's what happened, as near as I can make out: Sometime during the night there was a power failure affecting at least the part of the building where the MAE is located. This went undetected long enough for the UPS powering one of the Gigaswitches to drain, so the Gigaswitch lost power. When power was restored, the Gigaswitch refused to come back up, claiming that it's management memory was corrupted. We were finally able to clear the management memory and rebuild the Gigaswitch configuration from scratch, and it finally came back to life. For the short term, we're going to put that switch on the main DC power plant (as all the others are) which has a much longer lifetime and is watched better. We'll also try to find out what happened with the Gigaswitch management memory. Steve

1 0

Re: Unacceptable Delay
by Bill Woodcock 11 Jul '97

11 Jul '97

I would point out that the delay inherent in the NANOG list which keeps it from being useful to ops folks in real-time is exactly what spurred creation of the IPNmoo and the Inter-Provider Notification Project in general. Nobody who was logged into IPNmoo appears to have lost their connections to it. -Bill ______________________________________________________________________________ bill woodcock woody(a)zocalo.net woody(a)nowhere.loopback.edu user(a)host.domain.com

1 0