Survey on IBGP persistent route oscillation problem

Fri Mar 22 01:05:19 UTC 2002

Hi Jac,

I try to find a suitable bgp mib/oid for nearly one week, but failed. I have
to say bgp v4
mib provide little help on this issue, neither cisco proprietary mib. So I
have to turn to
the "primitive" method.

I use nc batch file ( netcat 1.1 for NT running in dos shell ) to read the
"sh ip b s | inc table"
output from each router, then a homemade perl script to analysis the router
output and extract
the table counter value, at last  the perl script was called by mrtg as an
external script.  All the
 three parts are very small, and working well together for months.

I've added some download link in the page (http://performance.cn.net:2003/),
you can have
them as your wish :-)

regards,

Yu Ning
______________________________________

(Mr.) Yu Ning, Chief Engineer
ChinaNET (AS4134) Senior Support
Internet Dep. DCBU, China Telecom
Beijing, P.R.China +86-10-62072357
Personal Page-> http://navidog.126.com
______________________________________

|-----Original Message-----
|From: Jac Kloots [mailto:Jac.Kloots at surfnet.nl]
|Sent: Thursday, March 21, 2002 10:08 PM
|To: Yu Ning
|Subject: RE: Survey on IBGP persistent route oscillation problem
|
|
|
|
|Yu Ning,
|
|I took a look at your page and was wondering how you read the BGP table
|version? do you use a snmpget command or do you have to login to
|the router?
|
|If you use SNMP for this value, which OID is it you're using?
|
|Regards,
|
|Jac
|
|> Hi,
|>
|> We have a similar situation (RR + always-compare-MED off), and
|the BGP table
|> version keeps changing at 1K/min (http://performance.cn.net:2003/). I
|> suspect some
|> route meet the criteria of IDR-oscillation draft. But in real world, it's
|> very hard to pick
|>  up the pattern depicted in the draft from a huge log of debug
|bgp output.
|>
|> After several sample from the huge bgp log, I have the following
|questions:
|>
|> 1.  How many time do our operator really find and affected by the problem
|> depicted
|>      in  draft-ietf-idr-route-oscillation-01.txt   £¿
|>
|> 2. From my experience, most flapping seems to be oscillation route which
|> escaped the
|>      eBGP damping protection. And for this, we need adjust damping
|> parameters from
|>      RIPE 220 to make up.
|>
|> 3.   Anyone see oscillation been magnified when injected into RR (cluster
|> structure) ?
|>
|>
|> 4. What's the typical bgp table change rate within your network ? From my
|> observation at
|>    the major global looking glass, should be  below 0.5K/min with some
|> accidental surge.
|>
|>
|>
|> regards,
|>
|> Yu Ning
|> ______________________________________
|>
|> (Mr.) Yu Ning, Chief Engineer
|> ChinaNET (AS4134) Senior Support
|> Internet Dep. DCBU, China Telecom
|> Beijing, P.R.China +86-10-62072357
|> Personal Page-> http://navidog.126.com
|> ______________________________________
|>
|> |-----Original Message-----
|> |From: owner-nanog at merit.edu [mailto:owner-nanog at merit.edu]On
|> |Behalf Of Susan Hares
|> |Sent: Wednesday, March 20, 2002 8:23 PM
|> |To: nanog at merit.edu
|> |Subject: Survey on IBGP persistent route oscillation problem
|> |
|> |
|> |
|> |
|> |
|> |IBGP and Persistent Route Oscillation
|> |
|> |The Draft BGP Persistent Route Oscillation draft describes route
|> |oscillation occurring in networks today.   This draft can be
|obtained at:
|> |
|>
||http://www.ietf.org/internet-drafts/draft-ietf-idr-route-oscillati
|on-01.txt
|> |
|> |In order to evaluate how significant a problem the route
|oscillation is, I
|> |would like to get input from network operators.   If you could
|take a few
|> |moments and fill out this survey, it would help in determining the
|> |severity
|> |of the problems.   Any information solicited by this survey will remain
|> |private.   Summaries of this information will be sent to this
|list and the
|> |IDR working list.
|> |
|> |
|> |Sue Hares
|> |IDR co-chair
|> |
|>
|>
|>
|
|--
|Jac Kloots
|Network Services
|SURFnet bv
|
|