Routing

Hi All,

 

I have this question: is it possible to configure (and have no problem) two or more VRF on the same PE with the same vrf-target import and export community?

 

Thank you

Antonio

 

 

Hello Antonio,

What are you trying to achieve exactly from a service perspective?

I suggest to have a look at this example:

http://www.juniper.net/techpubs/en_US/junos12.3/topics/example/multicast-vpn-mbgp-extranets-configuring-solutions.html

Thanks,
Ato

___________________________________
Accept as Solution = cool !
Accept as Solution+Kudo = You are a Star !

[paraphrasing master aarseniev]

Hi Ato,

 

I had this problem: I was trying to rename a vrf instance in order to obtain 3 other RI from the original one.I was trying to keep the same vrf-target import/export at each other but different RD. Also I'm using local-as loop and indipendent-domain knobs.

 

When I commited the configuration the RPD crashed. The router is a M320 with 10.4 release.

 

I hope I was clear.

 

Br

Antonio

 

Hi Antonio,

Sorry, it's unclear to me what you are trying to achieve from a service perspective. Why do you want 3 VRFs with the same RTs?

Ato

Hi Ato,

 

for a full-mesh connectivity. But focus on the question have you any idea about the RPD core?

 

 

Br

Antonio

Exotic configurations are more likely to trigger bugs. You can achieve full-mesh connectivity with a single VPN. And if you really want to have different VPNs, you can use extranet.

BTW you can import several targets in same VRF

Hi Antonio,

 

Can you send the output of:

 

show system core-dumps core-file-info <your_core_file>

 

Ato

HI Ato,

 

 

show system core-dumps core-file-info /var/tmp/rpd.core.1.gz

'rpd' process terminated with signal 6 Aborted.

Stack trace:
#0  0x88fd3f53 in kill () from /usr/lib/libc.so.6
#0  0x88fd3f53 in kill () from /usr/lib/libc.so.6
#1  0x88ed8371 in raise () from /usr/lib/libthr.so.2
#2  0x88fd2c7a in abort () from /usr/lib/libc.so.6
#3  0x88faccea in __assert () from /usr/lib/libc.so.6
#4  0x0862da6b in tag_dpvc_create_p2mp_pvc ()
#5  0x082bf77e in tai_create_dynamic_tunnel ()
#6  0x0879a816 in vpls_p2mp_router_id_changed ()
#7  0x0879a89d in vpls_p2mp_router_id_changed ()
#8  0x081d8892 in task_job_bg_dispatch ()
#9  0x081c966a in main ()

show system core-dumps core-file-info /var/tmp/rpd.core-tarball.1.tgz
'rpd' process terminated with signal 6 Aborted.

Stack trace:
#0  0x88fd3f53 in kill () from /usr/lib/libc.so.6
#0  0x88fd3f53 in kill () from /usr/lib/libc.so.6
#1  0x88ed8371 in raise () from /usr/lib/libthr.so.2
#2  0x88fd2c7a in abort () from /usr/lib/libc.so.6
#3  0x88faccea in __assert () from /usr/lib/libc.so.6
#4  0x0862da6b in tag_dpvc_create_p2mp_pvc ()
#5  0x082bf77e in tai_create_dynamic_tunnel ()
#6  0x0879a816 in vpls_p2mp_router_id_changed ()
#7  0x0879a89d in vpls_p2mp_router_id_changed ()
#8  0x081d8892 in task_job_bg_dispatch ()
#9  0x081c966a in main ()


Let me know if you find something

 

Br

Antonio

Doesn't match any known PR. What is the configuration change triggering the issue? Please attach complete config (as a file) and show | compare , before commit.

Hi Ato

 

I've attached a system compare.

 

Thank you

 

Antonio

I realize now that I've done some mistake during the changing of configuration.

 

Maybe that mistake caused a RPD core 😞

 

 

Antonio

 

Looking at the core dump stack trace, and at the configuration change, it seems to be related to router ID changes. So a one-time core dump.

 

___________________________________
Accept as Solution = cool !
Accept as Solution+Kudo = You are a Star !
/paraphrasing master Alex