Routing
Reply
Contributor
sunfun
Posts: 10
Registered: ‎09-04-2008
0

LDP down when CPU grows up to 100 percent

Hi,

 

We have a very strange problem on a M20 router:


Model: m20
JUNOS Base OS boot [8.2R4.5]
JUNOS Base OS Software Suite [8.2R4.5]
JUNOS Kernel Software Suite [8.2R4.5]
JUNOS Crypto Software Suite [8.2R4.5]
JUNOS Packet Forwarding Engine Support (M/T Common) [8.2R4.5]
JUNOS Packet Forwarding Engine Support (M20/M40) [8.2R4.5]
JUNOS Online Documentation [8.2R4.5]
JUNOS Routing Software Suite [8.2R4.5]

 

We always lose LDP peers, for a reason we haven't the correct solution to solve the problem.

That's why we've decided to activate a LDP debug, to find more details.

 

First we have a keepalive timer expired, and then, there's no hello adjacency for LDP peer.

And all LDP peers on this router fall down, whereas on another M20 router with the same JUNOS, we don't see this same kind of problem.

 

Let's see below the logs about this problem:

 

Sep 11 17:19:29.864210 LDP: Notification with error status, from 212.129.9.20
Sep 11 17:19:29.865088 RPD_LDP_SESSIONDOWN: LDP session 212.129.9.20 is down, reason: received notification from peer
Sep 11 17:19:29.867505 LDP rcvd TCP PDU 212.129.9.125 -> 212.129.9.8 (none)
Sep 11 17:19:29.867549 ver 1, pkt len 32, PDU len 28, ID 212.129.9.125:0
Sep 11 17:19:29.867575   Msg Notification (0x1), len 18, ID 7232783
Sep 11 17:19:29.867598     TLV Status (0x300), len 10
Sep 11 17:19:29.867622       Code 'Keepalive timer expired' (0x80000014)
Sep 11 17:19:29.867644       msgid 0, msgtype Unknown (0x0)


Sep 11 17:19:30.418175 RPD_LDP_SESSIONDOWN: LDP session 212.129.9.41 is down, reason: received notification from peer
Sep 11 17:19:32.169091 LDP LSP to 212.129.9.100 is down, reason: neighbor down.
Sep 11 17:19:32.169602 LDP LSP to 195.154.173.156 is down, reason: neighbor down.
Sep 11 17:19:32.217874 LDP rcvd TCP PDU 212.129.9.125 -> 212.129.9.8 (none)
Sep 11 17:19:32.217944 ver 1, pkt len 52, PDU len 48, ID 212.129.9.125:0
Sep 11 17:19:32.217970   Msg Initialization (0x200), len 38, ID 7232862
Sep 11 17:19:32.217993     TLV SesParms (0x500), len 14
Sep 11 17:19:32.218017     TLV GracefulRestartParms (0x8503), len 12
Sep 11 17:19:32.218052 LDP: Init to bound session from 212.129.9.125
Sep 11 17:19:32.218139 LDP sent TCP PDU 212.129.9.8 -> 212.129.9.125 (none)
Sep 11 17:19:32.218167 ver 1, pkt len 32, PDU len 28, ID 212.129.9.8:0
Sep 11 17:19:32.218190   Msg Notification (0x1), len 18, ID 38816580
Sep 11 17:19:32.218213     TLV Status (0x300), len 10
Sep 11 17:19:32.218236       Code 'No Hello adjacency' (0x80000010)
Sep 11 17:19:32.218259       msgid 0, msgtype Unknown (0x0)
Sep 11 17:19:32.220468 LDP rcvd TCP PDU 212.129.9.58 -> 212.129.9.8 (none)
Sep 11 17:19:32.220538 ver 1, pkt len 52, PDU len 48, ID 212.129.9.58:0
Sep 11 17:19:32.220563   Msg Initialization (0x200), len 38, ID 47925196
Sep 11 17:19:32.220586     TLV SesParms (0x500), len 14
Sep 11 17:19:32.220610     TLV GracefulRestartParms (0x8503), len 12


Sep 11 17:19:32.225343 LDP sent TCP PDU 212.129.9.8 -> 212.129.9.41 (none)
Sep 11 17:19:32.225369 ver 1, pkt len 32, PDU len 28, ID 212.129.9.8:0
Sep 11 17:19:32.225392   Msg Notification (0x1), len 18, ID 38816611
Sep 11 17:19:32.225414     TLV Status (0x300), len 10
Sep 11 17:19:32.225437       Code 'No Hello adjacency' (0x80000010)
Sep 11 17:19:32.225459       msgid 0, msgtype Unknown (0x0)
Sep 11 17:19:32.331823 LDP rcvd TCP PDU 212.129.9.100 -> 212.129.9.8 (none)
Sep 11 17:19:32.331901 ver 1, pkt len 36, PDU len 32, ID 212.129.9.100:0
Sep 11 17:19:32.331927   Msg Initialization (0x200), len 22, ID 6479407
Sep 11 17:19:32.331950     TLV SesParms (0x500), len 14
Sep 11 17:19:33.043149 LDP LSP to 212.129.9.208 is down, reason: topology changed.
Sep 11 17:19:33.043804 LDP LSP to 212.129.9.209 is down, reason: topology changed.
Sep 11 17:19:33.139111 LDP LSP to 212.129.9.117 is down, reason: topology changed.
Sep 11 17:19:33.139215 LDP LSP to 212.129.9.118 is down, reason: topology changed.
Sep 11 17:19:33.139255 LDP LSP to 212.129.14.8 is down, reason: topology changed.

 

Thank you for your help.

 

Best regards,

 

sunfun

Trusted Contributor
Samuel
Posts: 83
Registered: ‎11-15-2007
0

Re: LDP down when CPU grows up to 100 percent

Are you using BFD?
Contributor
sunfun
Posts: 10
Registered: ‎09-04-2008
0

Re: LDP down when CPU grows up to 100 percent

Hi Samuel,

 

As you can see after:

 

swo@router> show configuration | match bfd

{master}
swo@router>

 

No Bidirectional Forwarding Detection is used.

Anyway, this function concerns routing protocols, doesn't it?

 

I don't see any equivalence for LDP.

 

Thank you for your help.

 

Best regards,

 

sunfun

Trusted Contributor
Samuel
Posts: 83
Registered: ‎11-15-2007
0

Re: LDP down when CPU grows up to 100 percent

In fact we have already got some problem with BFD when the CPU grows up to 100%.

 

What about your IGP (OSPF / ISIS)? Do you have some logs?

Contributor
sunfun
Posts: 10
Registered: ‎09-04-2008
0

Re: LDP down when CPU grows up to 100 percent

Hi Samuel,

 

I don't see any strange logs about ospf and isis

First because ospf protocol is inactive on the router

 

##
## inactive: protocols ospf
##

 

Second, because there's nothing strange on isis logfile. I've activated flag about lsp-generation, ldp-synchronization and so on. Let's see the configuration i set on this router:

 

swo@router> show configuration protocols isis
traceoptions {
    file trace-isis size 1m files 10;
    flag error send receive;
    flag graceful-restart;
    flag ldp-synchronization;
    flag lsp-generation;
}

 

Let's see the logs about this configuration:

 

Sep 12 13:41:22.389559 Scheduling full LSP generation: reconfig
Sep 12 13:41:22.653719 Generating LSPs for L1
Sep 12 13:41:22.653988 Scheduling rebuild for L1 fragment BB1.TLH.00-00, sequence 0x315f8 in 0.020000s
Sep 12 13:41:22.654147 Add TE interface ge-0/0/0.0 to L1 LSP, reuse
Sep 12 13:41:22.654193 Add TE interface so-0/2/1.0 to L1 LSP, reuse
Sep 12 13:41:22.654232 Add TE interface so-1/0/0.0 to L1 LSP, reuse
Sep 12 13:41:22.654369 Generating LSPs for L2
Sep 12 13:41:22.654551 Scheduling rebuild for L2 fragment BB1.TLH.00-00, sequence 0x37abd in 0.020000s
Sep 12 13:41:22.654628 Scheduling rebuild for L2 fragment BB1.TLH.00-01, sequence 0x5d87 in 0.020000s
Sep 12 13:41:22.654697 Scheduling rebuild for L2 fragment BB1.TLH.00-02, sequence 0x43f6 in 0.020000s
Sep 12 13:41:22.654749 Scheduling rebuild for L2 fragment BB1.TLH.00-03, sequence 0x4e9d in 0.020000s
Sep 12 13:41:22.654805 Scheduling rebuild for L2 fragment BB1.TLH.00-04, sequence 0x4120 in 0.020000s
Sep 12 13:41:22.654860 Add TE interface ge-0/0/0.0 to L2 LSP, reuse
Sep 12 13:41:22.654899 Add TE interface so-1/0/0.0 to L2 LSP, reuse
Sep 12 13:41:22.657428 LSP generation complete, adding routes
Sep 12 13:41:22.701953 Rebuilding L1 fragment BB1.TLH.00-00, sequence 0x315f8
Sep 12 13:41:22.702153 Rebuilt L1 fragment BB1.TLH.00-00, size 317
Sep 12 13:41:22.702708 Rebuilding L2 fragment BB1.TLH.00-00, sequence 0x37abd
Sep 12 13:41:22.702858 Rebuilt L2 fragment BB1.TLH.00-00, size 1345
Sep 12 13:41:22.704031 Rebuilding L2 fragment BB1.TLH.00-01, sequence 0x5d87
Sep 12 13:41:22.704154 Rebuilt L2 fragment BB1.TLH.00-01, size 1250
Sep 12 13:41:22.705210 Rebuilding L2 fragment BB1.TLH.00-02, sequence 0x43f6
Sep 12 13:41:22.705323 Rebuilt L2 fragment BB1.TLH.00-02, size 1172
Sep 12 13:41:22.706331 Rebuilding L2 fragment BB1.TLH.00-03, sequence 0x4e9d
Sep 12 13:41:22.706453 Rebuilt L2 fragment BB1.TLH.00-03, size 1289
Sep 12 13:41:22.707484 Rebuilding L2 fragment BB1.TLH.00-04, sequence 0x4120
Sep 12 13:41:22.707679 Rebuilt L2 fragment BB1.TLH.00-04, size 1154
Sep 12 13:41:24.042459 Scheduling rebuild for L1 fragment BB1.TLH.00-00, sequence 0x315f9 in 0.020000s
Sep 12 13:41:24.042591 Scheduling rebuild for L2 fragment BB1.TLH.00-00, sequence 0x37abe in 0.020000s
Sep 12 13:41:24.152164 Rebuilding L1 fragment BB1.TLH.00-00, sequence 0x315f9
Sep 12 13:41:24.152349 Rebuilt L1 fragment BB1.TLH.00-00, size 317
Sep 12 13:41:24.153041 Rebuilding L2 fragment BB1.TLH.00-00, sequence 0x37abe
Sep 12 13:41:24.153200 Rebuilt L2 fragment BB1.TLH.00-00, size 1345
Sep 12 13:41:24.846743 Scheduling rebuild for L1 fragment BB1.TLH.00-00, sequence 0x315fa in 0.020000s
Sep 12 13:41:24.846859 Scheduling rebuild for L2 fragment BB1.TLH.00-00, sequence 0x37abf in 0.020000s
Sep 12 13:41:24.949481 Rebuilding L1 fragment BB1.TLH.00-00, sequence 0x315fa
Sep 12 13:41:24.949647 Rebuilt L1 fragment BB1.TLH.00-00, size 317
Sep 12 13:41:24.950463 Rebuilding L2 fragment BB1.TLH.00-00, sequence 0x37abf
Sep 12 13:41:24.950623 Rebuilt L2 fragment BB1.TLH.00-00, size 1345
Sep 12 13:41:25.421315 Scheduling rebuild for L1 fragment BB1.TLH.00-00, sequence 0x315fb in 4.529133s
Sep 12 13:41:30.249566 Rebuilding L1 fragment BB1.TLH.00-00, sequence 0x315fb
Sep 12 13:41:30.249746 Rebuilt L1 fragment BB1.TLH.00-00, size 317

 

thank you for your help

 

best regards,

 

sunfun

Contributor
xuzhengwei
Posts: 13
Registered: ‎09-04-2008
0

Re: LDP down when CPU grows up to 100 percent

Please paste the configurations of the M20 and it's peers' ldp.
Contributor
sunfun
Posts: 10
Registered: ‎09-04-2008
0

Re: LDP down when CPU grows up to 100 percent

Hi,

 

Let's see the configuration about LDP peers

 

swo@BB1.TLH> show configuration protocols ldp
traceoptions {
    file debug-ldp size 10m files 10;
    flag notification send receive detail;
    flag initialization send receive;
    flag state send receive;
    flag error send receive;
    flag event send receive;
}
track-igp-metric;
transport-address router-id;
inactive: interface ge-0/0/0.0;
inactive: interface so-0/2/0.0;
interface so-0/2/1.0;
inactive: interface so-1/0/0.0;
inactive: interface ge-1/2/0.0;
interface lo0.0;
session 212.129.9.1 {
    authentication-key "$9$WDlXVY24Zk.5bsfz69OB8X7VYoUDk"; ## SECRET-DATA
}
session 212.129.9.2 {
    authentication-key "$9$WDlXVY24Zk.5bsfz69OB8X7VYoUDk"; ## SECRET-DATA
}
session 212.129.9.3 {
    authentication-key "$9$WDlXVY24Zk.5bsfz69OB8X7VYoUDk"; ## SECRET-DATA
}
...

...

...

 

and so on.

 

 I can't put the whole configuration today.

Please let me know what the most important things you need.

 

Thank you for your help.

 

Best regards,

 

sunfun

Copyright© 1999-2013 Juniper Networks, Inc. All rights reserved.