Re: [mpls-linux-general] ldp_linux operation mode problems
Status: Beta
Brought to you by:
jleu
From: James R. L. <jl...@mi...> - 2002-06-19 05:18:30
|
Interesting. After I have chance to finish the ldp-portable/zebra integration I will work on debugging the differnt modes of operation. Jim On Tue, Jun 11, 2002 at 08:51:48PM +0200, Igor Vukomanovic wrote: > Hello, > > I've tried out several modes of operation for ldp-linux and I'm having some > rather problematic results. > The setup I used was: > > LER1---------------------LSR1---------------------LER2 > 192.168.20.200 192.168.20.201 192.168.19.206 > 192.168.19.204 > > All machines mpls-linux 0.996 patched kernel 2.4.12, ldp-portable CVS (a few > weeks ago) > > Firstly, when I changed from Downstream Unsolicited to Downstream on Demand > mode, after adding global and interface, ldp crashes. I browsed through > mailing list archives and found Amish Verna's message from 2002-04-11, the > same thing happened to him but there was no reply. Is there a solution to > this problem? > > Another thing I tried was testing whether ldp really works in Ordered > control mode, as the default is Ordered, but I found that it is not so. It > behaves as if it works in Independent mode! Let me explain my test: > In the setup above, the routing was: > LER1: > route add -host 192.168.20.201 gw 192.168.20.201 > route add -host 192.168.19.203 gw 192.168.20.201 > > LSR1: > route add 192.168.20.200 gw 192.168.20.200 > route add 192.168.19.206 gw 192.168.19.206 > route add 192.168.19.203 gw 192.168.19.206 > > LER2: > route add 192.168.19.204 gw 192.168.19.204 > > The LDP session was first setup between LSR1 and LER2. After that the LDP > session was setup between LSR1 and LER1. > > Note that 192.168.19.203 is a *fictional* FEC for which the LER1, when > working in the Independent mode, should get a label for, from the LSR1. > But when working in the Ordered mode, the LSR1 *must not* send any labels to > LER1 for a FEC 192.168.19.203 since LER2, the next hop LSR for that FEC > (from LSR1's point of view) didn't yet send labels for that FEC. (It didn't > send and it won't send because I never added that route in the routing > table, on purpose). > But, I noticed a label mapping message originated from LSR1 to LER1 with a > label for a FEC 192.168.19.203. > (I grabbed it with Ethereal, and can present it if neccessary). > > The result on LER1: > > voip5:/home/igor# cat /proc/net/mpls_fec > 40005403 192.168.19.203/32 > 40006003 192.168.20.201/32 > > voip5:/home/igor# cat /proc/net/mpls_out > 40005403 0/0/0 PUSH(gen 21) SET(lec0,192.168.20.201) > 40006003 102/6222/0 PUSH(gen 24) SET(lec0,192.168.20.201) > > According to RFC3036: > > " When using LSP ordered control, an LSR may initiate the transmission > of a label mapping only for a FEC for which it has a label mapping > for the FEC next hop, or for which the LSR is the egress. For each > FEC for which the LSR is not the egress and no mapping exists, the > LSR MUST wait until a label from a downstream LSR is received before > mapping the FEC and passing corresponding labels to upstream LSRs." > > Is LSR1 the Egress for a FEC 192.168.19.203? > > According to RFC3036: > > " An LSR may act as an egress LSR, with respect to a particular FEC, > under any of the following conditions: > 1. The FEC refers to the LSR itself (including one of its directly > attached interfaces). > 2. The next hop router for the FEC is outside of the Label > Switching Network. > 3. FEC elements are reachable by crossing a routing domain > boundary, such as another area for OSPF summary networks, or > another autonomous system for OSPF AS externals and BGP routes" > > Obviously none of the 1,2, or 3 applies. Therefore the LSR1 is not the > egress for that particular FEC, and it didn't yet receive the mapping from > the FEC next hop. It shouldn't have sent the label mapping message. > > According to the RFC, the fact that LSR1 works in Unsolicited label > distribution mode, doesn't change above facts; It only means the LSR1 is > responsible for sending labels upstream to the LER1, (doesn't wait for the > label request from LER1); In Downstream on Demand mode, the LER1 would be > responsible for those labels (by sending label request messages to LSR1). > > Sincerely, > > Igor Vukomanovic > > P.S. > Some additional info : > On LSR1, the result was also OK : > > voip4:/# cat /proc/net/mpls_out > 40004402 72/4560/0 PUSH(gen 17) SET(eth0,192.168.19.206) > 40004803 38/2681/0 PUSH(gen 18) SET(lec0,192.168.20.200) > > voip4:/# cat /proc/net/mpls_fec > 40004402 192.168.19.206/32 > 40004803 192.168.20.200/32 > > However, on LER2 there were no out labels or fec's. I guess this is the > known problem of one-way LSP Morvan Daniel Muller and some other people > encountered. However, if I add: > > route add -host 192.168.20.200 gw 192.168.19.204 > > on LER2, it would create an out label for the FEC 192.168.20.200 ! (still no > labels for 192.168.19.204 though). > > (Just a piece of curious information). > > > _______________________________________________________________ > > Multimillion Dollar Computer Inventory > Live Webcast Auctions Thru Aug. 2002 - http://www.cowanalexander.com/calendar > > > > _______________________________________________ > mpls-linux-general mailing list > mpl...@li... > https://lists.sourceforge.net/lists/listinfo/mpls-linux-general -- James R. Leu |