Re: [c-nsp] Ang: Making SUP720 cope better under BGP load

[Mattias Gyllenvarg]

I can vouch for the asr9k in regards too performance. But the software
still is not as stable as you might want.
On May 2, 2013 9:52 AM, gustav.ulan...@steria.se wrote:

 Hello Simon.
 We are using asr1k for peering purposes and Sup2T in the core. We also
 have some sup 720 as PE routers.
 We find that the ASR1001 is alot faster at establishing our BGP sessions
 than both sup 720 and 2T. I would look into the ASR9001. Seems to be much
 better box than an ASR 1k box when you spec it to be able to push around
 40G. Often turns out cheaper than ASR1k boxes also.

 Gustav Uhlander
 Communication  Infrastructure Engineer

 Steria AB
 Kungsbron 13
 Box 169
 SE-101 23 Stockholm
 Sweden

 Tel: +46 8 622 42 15
 Fax: +46 8 622 42 23
 Mobile: +46 70 962 71 03
 gustav.ulan...@steria.se
 www.steria.se


 -cisco-nsp-boun...@puck.nether.net skrev: -
 Till: cisco-nsp@puck.nether.net
 Från: Simon Lockhart **
 Sänt av: cisco-nsp-boun...@puck.nether.net
 Datum: 2012-12-07 14:29
 Ärende: [c-nsp] Making SUP720 cope better under BGP load

 All,

 I'm currently using SUP720-3BXL's in my BGP border devices.  Obviously the
 SUP720 is not a particularly fast CPU, so it is pretty slow at bringing up
 a
 lot of BGP sessions.

 On one particular box, I've got 250 BGP neighbours - 1 full table transit,
 2
 IGP to route-reflectors, and the rest are peering sessions at an IXP.
 Recently,
 the IXP did maintenance causing the interface to drop, and it bought the
 box to
 its knees. The BGP Router process takes all the available CPU while it
 tries
 to re-establish the BGP sessions. While this is happening, the SUP720
 seems to
 give up processing other stuff in a timely manner - and I see MPLS LDP
 drop,
 OSPF neighbours drop, and then BGP sessions drop due to hold timer expires.
 With all these drops, it causes even more CPU load, and the cycle
 continues.

 I've been talking to other SUP720 using ISPs, and it seems that some see
 this
 same effect, and others don't.

 Currently running 12.2(33)SXJ3

 Are there any tweaks that I can apply to the IOS config to make the SUP720
 cope better in this sort of situation? I'd be happy for the BGP sessions to
 take a lot longer to re-establish, if it didn't kill everything else in the
 process...

 And, as a follow-on question, given that the SUP720 is so under-powered for
 BGP, what other options do I have which would cope better? SUP-2T? Or, if
 I need to move away from the 6500, what's good for BGP routing with about
 20-40G of throughput (i.e. 4-8 * 10GE ports)? How does the ASR9k or ASR1k
 range fair for BGP performance?

 Many thanks in advance,

 Simon
 ___
 cisco-nsp mailing list  cisco-nsp@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@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@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/

[c-nsp] Ang: Making SUP720 cope better under BGP load

[Gustav . Ulander]

Hello Simon.We are using asr1k for peering purposes and Sup2T in the core. We also have some sup 720 as PE routers.We find that the ASR1001 is alot faster at establishing our BGP sessions than both sup 720 and 2T. I would look into the ASR9001. Seems to be much better box than an ASR 1k box when you spec it to be able to push around 40G. Often turns out cheaper than ASR1k boxes also.Gustav UhlanderCommunication  Infrastructure EngineerSteria ABKungsbron 13Box 169SE-101 23  StockholmSwedenTel: +46 8 622 42 15Fax: +46 8 622 42 23Mobile: +46 70 962 71 03gustav.ulan...@steria.se www.steria.se-cisco-nsp-boun...@puck.nether.net skrev: -Till: cisco-nsp@puck.nether.netFrån: Simon Lockhart Sänt av: cisco-nsp-boun...@puck.nether.netDatum: 2012-12-07 14:29Ärende: [c-nsp] Making SUP720 cope better under BGP loadAll,I'm currently using SUP720-3BXL's in my BGP border devices. Obviously theSUP720 is not a particularly fast CPU, so it is pretty slow at bringing up alot of BGP sessions.On one particular box, I've got 250 BGP neighbours - 1 full table transit, 2IGP to route-reflectors, and the rest are peering sessions at an IXP. Recently,the IXP did maintenance causing the interface to drop, and it bought the box toits knees. The "BGP Router" process takes all the available CPU while it triesto re-establish the BGP sessions. While this is happening, the SUP720 seems togive up processing other stuff in a timely manner - and I see MPLS LDP drop,OSPF neighbours drop, and then BGP sessions drop due to hold timer expires.With all these drops, it causes even more CPU load, and the cycle continues.I've been talking to other SUP720 using ISPs, and it seems that some see thissame effect, and others don't.Currently running 12.2(33)SXJ3Are there any tweaks that I can apply to the IOS config to make the SUP720cope better in this sort of situation? I'd be happy for the BGP sessions totake a lot longer to re-establish, if it didn't kill everything else in theprocess...And, as a follow-on question, given that the SUP720 is so under-powered forBGP, what other options do I have which would cope better? SUP-2T? Or, ifI need to move away from the 6500, what's good for BGP routing with about 20-40G of throughput (i.e. 4-8 * 10GE ports)? How does the ASR9k or ASR1krange fair for BGP performance?Many thanks in advance,Simon___cisco-nsp mailing list cisco-nsp@puck.nether.nethttps://puck.nether.net/mailman/listinfo/cisco-nsparchive at http://puck.nether.net/pipermail/cisco-nsp/___
cisco-nsp mailing list  cisco-nsp@puck.nether.net
https://puck.nether.net/mailman/listinfo/cisco-nsp
archive at http://puck.nether.net/pipermail/cisco-nsp/