Ask Your Question

jfarschman's profile - activity

2016-04-02 02:23:07 -0600 received badge  Famous Question (source)
2016-04-02 02:23:07 -0600 received badge  Notable Question (source)
2014-11-06 09:18:54 -0600 commented answer If I use nova-network, is floating IP the only way to go?

Sorry for the late comment, but yes, you guessed correctly. unless nova.fixed_ips had a reserved address that matches the address in your networking config it will break things.

2014-11-06 09:17:50 -0600 commented answer If I use nova-network, is floating IP the only way to go?

INSERT INTO fixed_ips (created_at, updated_at, deleted_at, id, address, network_id, allocated, leased, reserved, virtual_interface_id, host, instance_uuid, deleted) VALUES ('2013-08-06 21:15:54', '2013-08-06 21:15:54', NULL, 1030, '192.168.253.21', 1, 0, 0, 0, ........

2014-11-06 09:13:51 -0600 answered a question How to debug cluster where instance status is stuck on "scheduling"?

Morning,

I think Andy is correct when he says "Which is certainly RabbitMQ!". I faced a similar problem where a firewall would DROP packets without notifying the endpoints that the connections was dropped. This was done because the connection was idle for too long. Better behavior is to let both parties know the connection was dropped and should be reestablished. But, DROP, don't notify was the behavior.

So, my RabbitMQ connection was dropped every 2 hours. Restarting rabbit was all I needed to do to solve it.

My solution was to ask the FW team to change the behavior, but in the meantime, I also changed the sysctl.conf to have keepalive packets. Below is a little snippet from my puppet config.

"net.ipv4.tcp_keepalive_time" => { value => 60}, "net.ipv4.tcp_keepalive_intvl" => { value => 5}, "net.ipv4.tcp_keepalive_probes" => { value => 6},

Hope this helps.

2014-10-14 09:51:38 -0600 received badge  Teacher (source)
2014-10-14 09:34:19 -0600 answered a question If I use nova-network, is floating IP the only way to go?

Gators,

With nova-network you can make your FIXED space routable and just use that for everything. Ignore the FLOATING space altogether. I wasn't able to follow all the details of your setup, but if you provide nova-network with some IPs are routable on your network this will work. It's not how things are designed, but I've done this very successfully in a lab before. In nova.conf set things up to bridge

flat_network_bridge=br100
flat_interface=eth1
gateway=<computenodesip>

network_manager=nova.network.manager.FlatDHCPManager
force_dhcp_release=True
dhcpbridge_flagfile=/etc/nova/nova.conf
dhcpbridge=/usr/bin/nova-dhcpbridge
dhcp_domain=your.domain.com
dhcp_lease_time=1260
dnsmasq_config_file=/etc/nova/dnsmasq-nova.conf
auto_assign_floating_ip=false
use_single_default_gateway=false
libvirt_use_virtio_for_bridges=true
update_dns_entries=true

One problem you may have is that the FIXED network will try to take the entire space. For instance, if you build a FIXED network on 192.168.253.0/24 it will want to start allocating all of the IPs in this space. You can address this by removing individual addresses from nova.fixed_ips table.... you may also need to properly assign you compute nodes in the same table.

2014-08-13 06:51:48 -0600 received badge  Popular Question (source)
2013-11-11 11:15:39 -0600 answered a question Cant load myopenstack Dashboard...

Any time you create a keystone endpoint you also need to bring up the corresponding services in OpenStack. Failing to do that will result in an error just like the one we are seeing here. I just brought up a test box and tried to shortcut by not installing Cinder..... after already creating a cinder endpoint.

So, either remove the endpoint, or start the service.