Description

When ping the mgmt interface address from mgmt_junos instance, the ping is unsuccessfully and gets the error of "No such vrf (-JU) (ret:10)". This KB lists the info that helps in troubleshooting this kind of issue.

Symptoms

After upgrade JSU junos-evo-install-ptx-x86-64-22.4R2-S1-J1.3-EVO.iso, the mgmt interface can’t be used and failed to ping the mgmt interface address:

 

user@ptx10016> ping 192.168.3.65 routing-instance mgmt_junos count 5

 

No such vrf (-JU) (ret:10)

 

{master)

user@ptx10016>ping 192.168.3.66 routing-instance mgmt_junos

No such vrf (-JU) (ret: 10)

 

user@ptx10016> show interfaces terse routing-instance mgmt_junos

Interface              Admin Link Proto   Local                Remote

re0:mgmt-0.0           up    up   inet    192.168.3.66/26

                                  inet6   fdbd:xxxx:xxxx:x::xxxx:xxx/64

                                           fe80::xxxx:xxxx:xxxxx:xxx/64

re1:mgmt-0.0           up    up   inet     

                                   inet6   fe80::xxxx:xxxx: xxxx:xxx/64

 

 

user@ptx10016>show log messages| match vrf

 

 

Feb 2 21:18:05.505 2024 ptx10016-re1 login[31085]: %AUTHPRIV-2-PAM_TACPLUS_SEND_AUTHEN_FAIL: Failed in sending authentication with error Could not set socket options with vrf index: Protocol not available.

Feb 2 21:22:13.393 2024 ptx10016-re0 login[5514]: %AUTHPRIV-2-PAM_TACPLUS_SEND_AUTHEN_FAIL: Failed in sending authentication with error Could not set socket options with vrf index: Protocol not available.

Solution

Suggest to collect the below info to troubleshooting this issue:

 

start shell user root 

 

# ip addr show vrf mgmt_junos

# ifconfig vmb0

# echo $$

, get the id above

# getvrf <id>

# switchvrf $$ mgmt_junos

# getvrf <id>

# ping 10.52.3.66

# switchvrf $$ 

 

# ls -ltr /usr/evo/share

# cat /usr/evo/share/evostart_env

# nc -U /usr/evo/share/vrf_name.socket

# systemctl -l | egrep -i "vrf|mgmt"

# ip -d link show type vrf

# ip link show vrf mgmt_junos

# ip -br link show vrf mgmt_junos

 

get the journalctl log and the file /var/run/allfiles.usr_evo_share

# journalctl --since "1 day ago" > /var/tmp/journalctl_log_xx.log

 

Get the fibd trace and debug info:

> show trace node re0 application fibd

> request system debug-info mode lite node re0 since 1d

 

Suggest workaround:

request system application restart node re0 app fibd 

Modification History

2024-02-28 : Article Created