[SOLVED] frr.service doesn't restart when call ifreload -a ou doesn't start in boot time.

[Gilberto Ferreira]

Hi there.

I update a 3 node cluster with ceph from PVE 8 to PVE9.
On the PVE8 nodes it's has frr configuration, to provide mesh network between this 3 nodes.
After upgrade and restart, I notice that the networking hangs.
So I reboot and enter in rescue mode.
After commented the line:
#post-up systemctl restart frr.service
The service boot normally.

Now when need to use ifreload -a, the frr doesn't work, until I use systemctl restart frr manually.

Something happens between version 8 and 9 about frr service?

Thanks for you help.

 

[Gilberto Ferreira]

It's happens again...
New fresh installation.
After activated frr hungs on boot


I am using this options in daemon for frr

bgpd=yes
ospfd=yes
ospf6d=yes
ripd=no
ripngd=no
isisd=no
pimd=no
pim6d=no
ldpd=no
nhrpd=no
eigrpd=no
babeld=no
sharpd=no
pbrd=no
bfdd=yes
fabricd=yes
vrrpd=no
pathd=no

Any help, please?

 

[Gilberto Ferreira]

Hi... Do I need to something with fabric when upgrade from Proxmox 8 to Proxmox 9.
Everything was fine, but after upgrade to PVE 9, the post-up systemctl restart frr.services doesn't work inside /etc/network/interfaces.

 

[BobhWasatch]

You might be the only person running frr directly on their PVE host.

 

[Gilberto Ferreira]

You might be the only person running frr directly on their PVE host.

I'm not sure about that.
But I read the docs[0] again, and there's some additions that need in PVE 9.
I will check out.

SOLVED after change the line in the interfaces file:
Before:
vmbr0
...
...
post-up /usr/bin/systemctl restart frr.service
After:
vmbr0
...
...
post-up /usr/bin/systemctl is-active --quiet frr.service && /usr/bin/systemctl restart frr.service || true

[0] https://pve.proxmox.com/wiki/Full_Mesh_Network_for_Ceph_Server#/etc/network/interface_4