[Openstack] Folsom with Quantum on RHEL6.3
Hi I have a running Essex OpenStack cluster on Ubuntu with multi host network set up. Floating IP etc is working. I am planning to install same kind of setup with Folsom on RHEL 6.3 and quantum. I have some questions Most of the documents are mentioning quantum server with other nova services and a separate network server running dhcp agent and L2/L3 agents. Considering that I have a decent server, is it OK to merge all the services except compute node on same machine? If I want to support floating IP, do I have to install L3 agent on controller/network node? Is it necessary to have a separate data and management network ? I am not expecting a lot of data crossing the network initially. I am planning to use RHEL6.3, any thought about it. Thanks for any help. Cheers Kashif ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
Re: [Openstack] Folsom with Quantum on RHEL6.3
Hi Gary Thanks a lot for your detailed answer . Cheers Kashif On Thu, Jan 24, 2013 at 6:22 AM, Gary Kotton wrote: > On 01/24/2013 12:02 AM, mohammad kashif wrote: > > > Hi > > I have a running Essex OpenStack cluster on Ubuntu with multi host network > set up. Floating IP etc is working. I am planning to install same kind of > setup with Folsom on RHEL 6.3 and quantum. I have some questions > > Most of the documents are mentioning quantum server with other nova > services and a separate network server running dhcp agent and L2/L3 agents. > Considering that I have a decent server, is it OK to merge all the services > except compute node on same machine? > > > There are a number of different deployment options. The ideal setup would > be something like: > > http://docs.openstack.org/trunk/openstack-network/admin/content/connectivity.html > > > > If I want to support floating IP, do I have to install L3 agent on > controller/network node? > > > I would suggest having this on a network node. At the moment there is not > support to run multiple L3 agents on compute nodes. We are currently > working on a solution for this. Please note that there is no multi host > support for Quantum at the moment. > > > Is it necessary to have a separate data and management network ? I am not > expecting a lot of data crossing the network initially. > > > This is totally up to you. I would suggest having two different networks. > > > I am planning to use RHEL6.3, any thought about it. > > > RHEL6.3 could be supported by EPEL or through the preview at > redhat.com/openstack > > > Thanks for any help. > > Cheers > Kashif > > > > ___ > Mailing list: https://launchpad.net/~openstack > Post to : openstack@lists.launchpad.net > Unsubscribe : https://launchpad.net/~openstack > More help : https://help.launchpad.net/ListHelp > > > > ___ > Mailing list: https://launchpad.net/~openstack > Post to : openstack@lists.launchpad.net > Unsubscribe : https://launchpad.net/~openstack > More help : https://help.launchpad.net/ListHelp > > ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
[Openstack] openvswitch package for rhel6
Hi I am trying to install openstack on rhel6.4 beta and using epel repos. I am getting a dependency problem while installing openstack-quantum-openvswitch Package: openstack-quantum-openvswitch-2012.2.1-1.el6.noarch (epel) Requires: openvswitch I can't find openvswitch package anywhere. Can some one point this to me. Have any one manage to install openstack folsom with openvswitch on rhel ? Thanks Kashif ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
[Openstack] Nova network problem on rhel6.4 and folsom
Hi I am installing openstack folsom on rhel6.4 with multi_host nova network. I have a working setup with ubuntu 12.04 and Essex and I am using almost same network setup with rhel with folsom. I don't understand that what is going wrong with rhel. I have two machines with everything including nova-compute on one machine and an extra machine running nova-compute, nova-network and nova-metadata-api. Main problem is that I can create vm and it is getting private ip address from dhcp but that address is not being attached to bridge br100. Nova network is creating bridge but it stay in unknown state ip link show br100 7: br100: mtu 1500 qdisc noqueue state UNKNOWN link/ether 00:30:48:35:3d:ff brd ff:ff:ff:ff:ff:ff Some of network related setting is public_interface = eth0 flat_interface = eth1 fixed_range = 192.168.9.32/27 flat_network_dhcp_start = 192.168.9.33 connection_type = libvirt flat_injected = False multi_host = True flat_network_dns = flat_network_bridge = br100 first machine have eth0 = public ip eth1 = 192.168.9.1 second machine em1 = public ip em2 = 192.168.9.2 I am creating nova network like this nova-manage network create private --fixed_range_v4=192.168.9.32/27--num_network=1 --bridge=br100 --network_size=32 --multi_host=True Can some one suggest some thing. Thanks Kashif ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
Re: [Openstack] Nova network problem on rhel6.4 and folsom
Hi Vish Thanks for looking into it. For the simplicity, I have disabled second compute node for time being. I am checking it as single node installation. Like one VM is running on controller node aad411aa-dc50-4fad-ab2f-8fd283ed94e8 | | ACTIVE | private=192.168.9.34 brctl show br100 bridge name bridge id STP enabled interfaces br100 8000.003048353dff no eth1 vnet0 ip add 1: lo: mtu 16436 qdisc noqueue state UNKNOWN link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 127.0.0.1/8 scope host lo inet 169.254.169.254/32 scope link lo inet6 ::1/128 scope host valid_lft forever preferred_lft forever 2: eth0: mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:30:48:35:3d:fe brd ff:ff:ff:ff:ff:ff inet 163.1.5.243/24 brd 163.1.5.255 scope global eth0 inet6 fe80::230:48ff:fe35:3dfe/64 scope link valid_lft forever preferred_lft forever 3: eth1: mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:30:48:35:3d:ff brd ff:ff:ff:ff:ff:ff inet6 fe80::230:48ff:fe35:3dff/64 scope link valid_lft forever preferred_lft forever 4: virbr0: mtu 1500 qdisc noqueue state UNKNOWN link/ether 52:54:00:e4:6d:24 brd ff:ff:ff:ff:ff:ff inet 192.168.122.1/24 brd 192.168.122.255 scope global virbr0 5: virbr0-nic: mtu 1500 qdisc noop state DOWN qlen 500 link/ether 52:54:00:e4:6d:24 brd ff:ff:ff:ff:ff:ff 7: br100: mtu 1500 qdisc noqueue state UNKNOWN link/ether 00:30:48:35:3d:ff brd ff:ff:ff:ff:ff:ff inet 192.168.9.33/27 brd 192.168.9.63 scope global br100 inet 192.168.9.1/24 brd 192.168.9.255 scope global br100 inet6 fe80::8cc4:2cff:feb2:2233/64 scope link valid_lft forever preferred_lft forever 8: vnet0: mtu 1500 qdisc pfifo_fast state UNKNOWN qlen 500 link/ether fe:16:3e:5f:b2:0a brd ff:ff:ff:ff:ff:ff inet6 fe80::fc16:3eff:fe5f:b20a/64 scope link valid_lft forever preferred_lft forever I don't understand that why br100 is displaying unknown state. Thanks Kashif On Tue, Feb 26, 2013 at 7:19 PM, Vishvananda Ishaya wrote: > > On Feb 26, 2013, at 10:11 AM, mohammad kashif > wrote: > > Hi > I am installing openstack folsom on rhel6.4 with multi_host nova network. > I have a working setup with ubuntu 12.04 and Essex and I am using almost > same network setup with rhel with folsom. I don't understand that what is > going wrong with rhel. > > I have two machines with everything including nova-compute on one machine > and an extra machine running nova-compute, nova-network and > nova-metadata-api. > > Main problem is that I can create vm and it is getting private ip address > from dhcp but that address is not being attached to bridge br100. Nova > network is creating bridge but it stay in unknown state > > ip link show br100 > > 7: br100: mtu 1500 qdisc noqueue state > UNKNOWN > link/ether 00:30:48:35:3d:ff brd ff:ff:ff:ff:ff:ff > > Some of network related setting is > > public_interface = eth0 > flat_interface = eth1 > > > These needs to be set explicitly to em1 / em2 on the second machine. If > you do brctl show i suspect em2 is not being added to the bridge. You could > also manually add it to the bridge: > brctl addif br100 em2 > > Also the private ip address is not added to the bridge. if you do: > brctl show br100 > > you will see that each vm gets a vnet* device that is added to the bridge. > The ip only exists internally to the vm. > > fixed_range = 192.168.9.32/27 > flat_network_dhcp_start = 192.168.9.33 > connection_type = libvirt > flat_injected = False > multi_host = True > flat_network_dns = > flat_network_bridge = br100 > > first machine have > eth0 = public ip > eth1 = 192.168.9.1 > > second machine > em1 = public ip > em2 = 192.168.9.2 > > > I am creating nova network like this > nova-manage network create private > --fixed_range_v4=192.168.9.32/27--num_network=1 --bridge=br100 > --network_size=32 --multi_host=True > > > Can some one suggest some thing. > > Thanks > Kashif > > > ___ > Mailing list: https://launchpad.net/~openstack > Post to : openstack@lists.launchpad.net > Unsubscribe : https://launchpad.net/~openstack > More help : https://help.launchpad.net/ListHelp > > > ___ Mailing list: https://launchpad.net/~openstack Post to : openstack@lists.launchpad.net Unsubscribe : https://launchpad.net/~openstack More help : https://help.launchpad.net/ListHelp
[Openstack] Fwd: Nova network problem on rhel6.4 and folsom
Sorry I didn't reply to list at first instance so I am forwarding it now Hi Vish Did some poking around but not sure about my conclusion. If you look at ip link show eth1 3: eth1: mtu 1500 qdisc pfifo_fast state UP qlen 1000 link/ether 00:30:48:35:3d:ff brd ff:ff:ff:ff:ff:ff br100: mtu 1500 qdisc noqueue state UNKNOWN link/ether 00:30:48:35:3d:ff brd ff:ff:ff:ff:ff:ff brctl show br100 bridge name bridge id STP enabled interfaces br100 8000.003048353dff no eth1 If I compare it with my Ubuntu installation then eth1 should be showing br100 as its master. ubuntu example ip link show eth0 3: eth0: mtu 1500 qdisc mq master br100 state UP qlen 1000 link/ether 00:1a:a0:2e:ff:58 brd ff:ff:ff:ff:ff:ff I tried manually setting up br100 as master for rhel6.4 machine but it turn out that ip command available with rhel6.4 does not support bridge. ip -V ip utility, iproute2-ss091226 Which is quite old. As I said I could not reach on any conclusion as if it is the main reason then it would have been detected earlier. Thanks Kashif On Tue, Feb 26, 2013 at 10:49 PM, Vishvananda Ishaya wrote: > Odd, I couldn't say. Is there vm working? Perhaps it doesn't matter. It is > a bit odd that you have two overlapping ips on br100. If something is > broken, that could potentially be causing issues. > > Vish > > On Feb 26, 2013, at 2:33 PM, mohammad kashif > wrote: > > Hi Vish > Thanks for looking into it. For the simplicity, I have disabled second > compute node for time being. I am checking it as single node installation. > > Like one VM is running on controller node > aad411aa-dc50-4fad-ab2f-8fd283ed94e8 | | ACTIVE | > private=192.168.9.34 > > brctl show br100 > bridge name bridge id STP enabled interfaces > br100 8000.003048353dff no eth1 > vnet0 > > ip add > 1: lo: mtu 16436 qdisc noqueue state UNKNOWN > link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 > inet 127.0.0.1/8 scope host lo > inet 169.254.169.254/32 scope link lo > inet6 ::1/128 scope host >valid_lft forever preferred_lft forever > 2: eth0: mtu 1500 qdisc pfifo_fast state > UP qlen 1000 > link/ether 00:30:48:35:3d:fe brd ff:ff:ff:ff:ff:ff > inet 163.1.5.243/24 brd 163.1.5.255 scope global eth0 > inet6 fe80::230:48ff:fe35:3dfe/64 scope link >valid_lft forever preferred_lft forever > 3: eth1: mtu 1500 qdisc pfifo_fast state > UP qlen 1000 > link/ether 00:30:48:35:3d:ff brd ff:ff:ff:ff:ff:ff > inet6 fe80::230:48ff:fe35:3dff/64 scope link >valid_lft forever preferred_lft forever > 4: virbr0: mtu 1500 qdisc noqueue state > UNKNOWN > link/ether 52:54:00:e4:6d:24 brd ff:ff:ff:ff:ff:ff > inet 192.168.122.1/24 brd 192.168.122.255 scope global virbr0 > 5: virbr0-nic: mtu 1500 qdisc noop state DOWN qlen > 500 > link/ether 52:54:00:e4:6d:24 brd ff:ff:ff:ff:ff:ff > 7: br100: mtu 1500 qdisc noqueue state > UNKNOWN > link/ether 00:30:48:35:3d:ff brd ff:ff:ff:ff:ff:ff > inet 192.168.9.33/27 brd 192.168.9.63 scope global br100 > inet 192.168.9.1/24 brd 192.168.9.255 scope global br100 > inet6 fe80::8cc4:2cff:feb2:2233/64 scope link >valid_lft forever preferred_lft forever > 8: vnet0: mtu 1500 qdisc pfifo_fast > state UNKNOWN qlen 500 > link/ether fe:16:3e:5f:b2:0a brd ff:ff:ff:ff:ff:ff > inet6 fe80::fc16:3eff:fe5f:b20a/64 scope link >valid_lft forever preferred_lft forever > > I don't understand that why br100 is displaying unknown state. > > Thanks > Kashif > > > > > > On Tue, Feb 26, 2013 at 7:19 PM, Vishvananda Ishaya > wrote: > >> >> On Feb 26, 2013, at 10:11 AM, mohammad kashif >> wrote: >> >> Hi >> I am installing openstack folsom on rhel6.4 with multi_host nova network. >> I have a working setup with ubuntu 12.04 and Essex and I am using almost >> same network setup with rhel with folsom. I don't understand that what is >> going wrong with rhel. >> >> I have two machines with everything including nova-compute on one machine >> and an extra machine running nova-compute, nova-network and >> nova-metadata-api. >> >> Main problem is that I can create vm and it is getting private ip >> address from dhcp but that address is not being attached to bridge br100. >> Nova network is creating bridge but it stay in unknown state >> >> ip link show br100 >> >> 7: br100: mtu 1500 qdisc noqueue state >> UNKNOWN >> link/ether 00:30:48:35:3d:ff brd ff:ff:ff:ff:ff:ff >> >