[j-nsp] PROBLEMS WITH TRAFFIC FORWARDING EX 4200 IN VIRTUAL CHASSIS

Eugeniu Patrascu eugen at imacandi.net
Sat Apr 18 16:29:05 EDT 2015


The PFE forwards the packets, it doesn't matter if the switch is running
only Layer2 things.

If you run Layer2 only, then just enable non-stop-bridging so when things
go weird, the other switch will become active RE but without taking a hit
on the forwarding plane.

Regards,
Eugeniu

On Fri, Apr 17, 2015 at 9:11 PM, Mario Andres Rueda Jaimes <
maeve2009 at gmail.com> wrote:

> Hi mike,
>
> Thanks for your quick response, however my VC environment is running in
> pure Layer 2, do you think the effect of restarting the PFE will have the
> same effect as your layer 3 environment ?
>
> Thanks
>
> MR
>
> On Fri, Apr 17, 2015 at 12:27 PM, Mike Gonnason <gonnason at gmail.com>
> wrote:
>
> > You can try restarting the PFE process.
> >
> > I had to restart/kill the PFE process when I was configuring Filter Based
> > Forwarding to send traffic to a new instance. The PFE wouldn't install
> > routes into the new instance, even with a config force. This was 2 EX4200
> > in a VC under JUNOS 11.4R2.14.
> >
> > It can help reduce downtime if you can have HA features enabled.
> >
> >
> > root at switch> start shell
> >
> > root at switch:RE:0% ps -aux | grep pfe
> >
> > root  1284  1.0  2.3 86804 23388  ??  S     6:57AM   0:53.90
> /usr/sbin/pfem
> > –N                          <======== 1284 is the process id
> > root    16  0.0  0.0     0    16  ??  WL    6:55AM   0:00.00
> > [mpfe_drv_taskq16:
> > root    44  0.0  0.0     0    16  ??  IL    6:55AM   0:00.00
> > [if_pfe_listen]
> > root  1305  0.0  1.1 25656 10892  ??  S     6:57AM   0:00.95
> /usr/sbin/pfed
> > -N
> > root  1825  0.0  0.1  2388  1052  p0  S+    7:16AM   0:00.01 grep pfe
> >
> > root at switch:RE:0% kill -9 1284
> >
> >                 <======== restart the process
> >
> >
> > Note:- Restarting the pfe would lead to an outage of maximum 30 – 40
> > seconds. If GRES, NSB and NSR are enabled the outage will be reduced to
> > less than 10 seconds.
> >
> > root at switch# set routing-options nonstop-routing
> >       <--------Enable Nonstop active routing
> > root at switch# set ethernet-switching-options nonstop-bridging
> > <--------Enable Nonstop active bridging
> > root at switch# set chassis redundancy graceful-switchover
> > <--------Enable GRES
> >
> > -Mike Gonnason
> >
> >
> > On Fri, Apr 17, 2015 at 9:20 AM, Mario Andres Rueda Jaimes <
> > maeve2009 at gmail.com> wrote:
> >
> > > Hi All
> > >
> > >
> > > We are having a bizarre issue with a Virtual chassis with 2 ex 4200.
> > > running 10.3R3.7
> > >
> > > Suddenly some of the ports of member 0 stopped forwarding traffic with
> > > apparently no reason, no logs, no physical issues, even more mac
> learning
> > > process was functioning normally in the affected ports
> > >
> > > It looks like a PFE problem but unfortunenlly we can´t open a TT with
> > JTAC.
> > >
> > > Does anybody experienced this kind of issues before ?, we appreciate
> any
> > > input
> > >
> > >
> > > Thanks
> > >
> > > MR
> > > _______________________________________________
> > > juniper-nsp mailing list juniper-nsp at puck.nether.net
> > > https://puck.nether.net/mailman/listinfo/juniper-nsp
> > _______________________________________________
> > juniper-nsp mailing list juniper-nsp at puck.nether.net
> > https://puck.nether.net/mailman/listinfo/juniper-nsp
> _______________________________________________
> juniper-nsp mailing list juniper-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/juniper-nsp
>


More information about the juniper-nsp mailing list