Skip to Content.

rare-users - Re: [RARE-users] [freertr] Debug interface down DPDK implementation

Subject: RARE user and assistance email list

List archive


Re: [RARE-users] [freertr] Debug interface down DPDK implementation


Chronological Thread 
  • From: Xavier Jeannin <>
  • To:
  • Subject: Re: [RARE-users] [freertr] Debug interface down DPDK implementation
  • Date: Tue, 1 Mar 2022 17:33:15 +0100
  • Dkim-filter: OpenDKIM Filter v2.10.3 zmtaauth01.partage.renater.fr 5E8D71401E8

Thank you Csaba

Xavier

Le 01/03/2022 à 17:30, mc36 a écrit :
hi,

On 3/1/22 16:26, Xavier Jeannin wrote:
With my colleague Maxime, we finally get the interface up thanks to a cold reload (a warm reload was not efficient).

so if reload warm did not helped but reload cold did then the by best guss is the processes were stopped in systemd...

I would like to ask another question about interface configuration.
Our interface was 10Gb/s. At one point we suspected that the speed of the interface was well configured on the path.

We used the parameter "bandwidth". The show command write 8000kbps when it 10Gbps_
_If we set to 10Gbps one interface without configuring the other ( 8000kbps), it still works.
I think that the bw/speed was in fact configured correctly in the FreeRtr/RARE.
Is there a specific way to find what is the bw/speed of the interface? is it at Debian level?


so the bandwidth command on the sdn interface (much like in cisco) have nothing to do with interface speeds...
they're primarily used for qos shaper/policer when the class uses percentage instead of explicit bandwidth,
and in ospf/isis rsvp-te to advertise the remaining bandwidth for the link...
the speed of the interfaces are configured under server p4lang:
sid(cfg)#server p4lang p4
sid(cfg-server)#export-port sdn1 ?
  <num>   - port number
  dynamic - dynamic port number

sid(cfg-server)#export-port sdn1 1 ?
  <cr>
  <num> - speed in gbps

sid(cfg-server)#export-port sdn1 1 10 ?
  <cr>
  <num> - fec, see hw vendor manual

sid(cfg-server)#export-port sdn1 1 10 1 ?
  <cr>
  <num> - autoneg, see hw vendor manual

sid(cfg-server)#export-port sdn1 1 10 1 1 ?
  <cr>
  <num> - flowctrl, see hw vendor manual

sid(cfg-server)#export-port sdn1 1 10 1 1 1 ?
  <cr>

sid(cfg-server)#export-port sdn1 1 10 1 1 1


regrads,
cs


/DAV0102#show interfaces sdn5//
//sdn5 is up (since 00:51:32, 9 changes)//
//  description: 10G-FP2[eno7 - 05:00.1 - 18:5a:58:d8:d4:65]//
//  type is sdn, hwaddr=0014.0b0d.536c, mtu=1500, //_bw=8000kbps_//, vrf=inet//
//  ip4 address=11.11.11.4/24, netmask=255.255.255.0, ifcid=56988446//
//  received 10227 packets (680426 bytes) dropped 91 packets (7931 bytes)//
//  transmitted 10143 packets (674184 bytes) promisc=false macsec=false//
/

Thank you
Xavier do


Le 28/02/2022    15:03, Xavier Jeannin a   crit  :

Le 28/02/2022    14:52, mc36 a   crit  :
hmm, then as a last resort,
No problem, if reboot the machines did not work, I will wait Fr  d  ric
what about "reload cold" on both sides of the link?

OK let do this tomorrow

Thank you
Regards
Xavier


tomorrow i'll be in the kifu office... maybe a day after tomorrow?
thanks,
cs

On 2/28/22 14:23, Xavier Jeannin wrote:

Le 28/02/2022       11:06, mc36 a     crit   :
exactly the 0 changes is the lying here...
all the dataplanes start the ports in down, then an up....
imho bffwd is not connected at all....
isn't it fl's developer box by the way? :)
Yes DAV001 is Wedge and is the the developer box of Fr   d ric.

Fr   d   ric kindly put the machine in a state in which when you reboot it, one is not obliged to restart the different required elements one by one (things that I will not be able to do).

Do you think that the origin of the problem is on the DPDK machine (DAV0102 DELL VEP) where the port is done?

I will go to RENATER premises and look physically to the port tomorrow.
Would you be available fir a debug VC session tomorrow? there is no obligation as you are very busy and this could wait.

Thank you
Xavier

thanks,
cs

On 2/28/22 11:03, Xavier Jeannin wrote:
DAV0001#show interfaces sdn15
sdn15 is promisc, *up *(since 2d16h, 0 changes)

--
Xavier JEANNIN
Network Dpt.
GN4-3 Network technology evolution (WP6 T1) task leader
T  l. : +33 1 5394 2042
Mob. : +33 6 3017 0963
www.renater.fr

--
Xavier JEANNIN
Network Dpt.
GN4-3 Network technology evolution (WP6 T1) task leader
Tél. : +33 1 5394 2042
Mob. : +33 6 3017 0963
www.renater.fr




Archive powered by MHonArc 2.6.19.

Top of Page