Ask Your Question
0

Howto analyze package loss an ovs br-int br-ext?

asked 2019-05-07 10:27:56 -0500

BenvanDamme gravatar image

updated 2019-05-21 10:26:54 -0500

I am experiencing package loss on br-ex and br-int but i am unable to find the reason why. The package loss occurs on the port named "local" on both bridges.

(openvswitch-vswitchd)[root@DIR3 /]# ovs-ofctl dump-ports br-ex
 OFPST_PORT reply (xid=0x2): 3 ports
   port LOCAL: rx pkts=0, bytes=0, drop=1182800, errs=0, frame=0, over=0, crc=0
           tx pkts=0, bytes=0, drop=0, errs=0, coll=0
  port  ens2f1: rx pkts=8700125, bytes=3802845819, drop=14, errs=0, frame=0, over=0, crc=0
           tx pkts=18963759, bytes=19909567218, drop=0, errs=0, coll=0
  port  "phy-br-ex": rx pkts=9059475, bytes=19745048191, drop=?, errs=?, frame=?, over=?, crc=?
           tx pkts=7590224, bytes=3669691741, drop=?, errs=?, coll=?

ovs-ofctl dump-ports br-int
OFPST_PORT reply (xid=0x2): 22 ports
  port "ha-5172b4d6-02": rx pkts=273625, bytes=10934496, drop=38, errs=0, frame=0, over=0, crc=0
           tx pkts=14, bytes=972, drop=0, errs=0, coll=0
  port  "ha-c31af944-ab": rx pkts=219479, bytes=8769532, drop=4, errs=0, frame=0, over=0, crc=0
           tx pkts=54199, bytes=2926782, drop=0, errs=0, coll=0
  port LOCAL: rx pkts=0, bytes=0, drop=1481594, errs=0, frame=0, over=0, crc=0
           tx pkts=0, bytes=0, drop=0, errs=0, coll=0

I would really appreciate any hint

edit retag flag offensive close merge delete

Comments

I have never tried it, but openvswitch seems to have a tracing tool.

Bernd Bausch gravatar imageBernd Bausch ( 2019-05-07 19:32:35 -0500 )edit

Thank you for your answer. Did not know that one. But it is not a general routing problem. Sometimes i lose some packets every few seconds and sometimes it runs fine.

BenvanDamme gravatar imageBenvanDamme ( 2019-05-08 02:27:33 -0500 )edit

Intermittent package loss? Is there another device or interface in the network with the same IP or MAC address?

Bernd Bausch gravatar imageBernd Bausch ( 2019-05-08 05:08:17 -0500 )edit

There are no duplicate IPs oder MAC-Addresses in the os environment.

64 bytes from 192.168.13.35: icmp_seq=1365 ttl=63 time=1.103 ms
Request timeout for icmp_seq 1366
Request timeout for icmp_seq 1367
64 bytes from 192.168.13.35: icmp_seq=1368 ttl=63 time=1.589 ms
BenvanDamme gravatar imageBenvanDamme ( 2019-05-08 08:04:56 -0500 )edit

There is sometimes enough loss for ssh-connections to reset for example.

BenvanDamme gravatar imageBenvanDamme ( 2019-05-08 08:05:28 -0500 )edit

1 answer

Sort by ยป oldest newest most voted
0

answered 2019-05-22 16:01:54 -0500

chalans gravatar image

Hello,

Maybe you can try to configure sflow on each OVS and send to a collector For our tests we used sFlowTrend (free version limited to 5 device)

You juste need to configure each OVS with this command :

docker exec -it -u 0 neutron_openvswitch_agent ovs-vsctl --id=@sflow create sflow agent=bond_pub.XXX target="\"XXX:6343\"" header=128 sampling=64 polling=10 -- set bridge br-tun sflow=@sflow

edit flag offensive delete link more

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower

Stats

Asked: 2019-05-07 10:27:56 -0500

Seen: 27 times

Last updated: May 22