Ask Your Question
0

dhcp works but not stable

asked 2015-07-09 03:10:32 -0500

vladber gravatar image

updated 2015-07-09 03:14:39 -0500

Hi ,

I have a Juno HA deployed using Mirantis 6.1 and having the following problem - if I launch a single VM it gets IP from dhcp but when I run multiple VMs simultaneously only 50% of them succeed to get IP .

Any ideas ?

/Vlad.

edit retag flag offensive close merge delete

Comments

All instances on same compute node?

capsali gravatar imagecapsali ( 2015-07-09 11:47:00 -0500 )edit

1 answer

Sort by ยป oldest newest most voted
0

answered 2015-07-09 03:19:36 -0500

uts9 gravatar image

Can you check for DHCP timeouts?

#tcpdump -i <interface> port 67 or port 68 -e -n
edit flag offensive delete link more

Comments

I get this output

# tcpdump -i any  port 67 or port 68 -e -n



08:36:58.845871   B fa:16:3e:e4:bb:ab ethertype 802.1Q (0x8100), length 338: vlan 1018, p 0, ethertype IPv4, 0.0.0.0.68 > 255.255.255.255.67: BOOTP/DHCP, Request from fa:16:3e:e4:bb:ab, length 290
08:36:58.845918 Out fa:16:3e:e4:bb:a
vladber gravatar imagevladber ( 2015-07-09 03:38:37 -0500 )edit

There should be a " BOOTP/DHCP, Reply" message from DHCP Server.

Please follow the steps provided in this article. Section: Debugging DHCP Issues with nova-network

uts9 gravatar imageuts9 ( 2015-07-09 03:56:17 -0500 )edit

The point is that a single VM does get IP . The problem starts when I run multiple VMs - only 50 % succeed to get IP .

vladber gravatar imagevladber ( 2015-07-09 05:22:03 -0500 )edit

Get to know Ask OpenStack

Resources for moderators

Question Tools

1 follower

Stats

Asked: 2015-07-09 03:10:32 -0500

Seen: 134 times

Last updated: Jul 09 '15