Brenne's profile - activity

You may want to look at Ironic. In this tutorial nested VMs are created and deployed via TFTP and I also suppose PXE.

Already heared about Ironic? Despite it is supposed to be used for baremetal deployment you can use it to start nested VMs described in this tutorial

have a look at the RDO Installation errors page. Your problem seems similar to the one described under the heading "Error creating database"

Hi,

have you used devstack to install OpenStack? Since you're writing about local.conf I assume you used it. Insert the following lines in your local.conf

[[local|localrc]]
disable_service n-net
enable_service q-svc
enable_service q-agt
enable_service q-dhcp
enable_service q-l3
enable_service q-meta
enable_service neutron
# Optional, to enable tempest configuration as part of devstack
enable_service tempest

Have a look at NeutronDevstack page for more information.

it's only a guess but you might want to try it, Enable IP forwarding on the compute node echo "1" > /proc/sys/net/ipv4/ip_forward

Hi,

how can I specify the interval in which the (free) resources of my compute node are updated? I looked into nova.conf and already changed the following parameters:

heal_instance_info_cache_interval = 20
service_down_time = 20
periodic_fuzzy_delay = 20
qpid_heartbeat = 20
scheduler_driver_task_period = 20

call_timeout = 20
db_check_interval = 20

But the inteval in which my compute node resources are updated still seems to be one minute. Which parameter in nova.conf should I change? Or is this configuration done somewhere else?

Hi Daniel,

you have to setup networking on the compute node and also copy the nova.conf from your controller node. Have you already looked into adding additional compute nodes?

Hi,

I'm working with devstack and ironic and I'm trying to do the steps in Deploying Ironic with Devstack. Everthing works fine until the command:

nova boot --flavor baremetal --image $image --key-name default testing

My ironic node is powered on but remains in provisioning state "wait call-back". The instance stucks in spwaning state. I went through the logs but couldn't find any error. Even virsh list --all shows that the baremetal node is powered on.

With tcpdump I see the boot requests of the (nested) baremetal instance. Also dnsmasq is running:

ps -aux |grep dnsmasq
nobody   32163  0.0  0.0  28204  1060 ?        S    Jun12   0:00 dnsmasq --no-hosts --no-resolv --strict-order --bind-interfaces --interface=tap6954347a-fd --except-interface=lo --pid-file=/opt/stack/data/neutron/dhcp/dc205004-4726-4a47-a458-09266cbbda3c/pid --dhcp-hostsfile=/opt/stack/data/neutron/dhcp/dc205004-4726-4a47-a458-09266cbbda3c/host --addn-hosts=/opt/stack/data/neutron/dhcp/dc205004-4726-4a47-a458-09266cbbda3c/addn_hosts --dhcp-optsfile=/opt/stack/data/neutron/dhcp/dc205004-4726-4a47-a458-09266cbbda3c/opts --leasefile-ro --dhcp-range=set:tag0,10.1.0.0,static,86400s --dhcp-lease-max=256 --conf-file= --domain=openstacklocal

But there seems to be no reply to the DHCP boot requests:

sudo tcpdump -i brbm
tcpdump: WARNING: brbm: no IPv4 address assigned
tcpdump: verbose output suppressed, use -v or -vv for full protocol decode
listening on brbm, link-type EN10MB (Ethernet), capture size 65535 bytes
09:18:32.110281 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP, Request from 52:54:00:48:6d:27 (oui Unknown), length 398
09:18:33.096545 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP, Request from 52:54:00:48:6d:27 (oui Unknown), length 398
09:18:35.073842 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP, Request from 52:54:00:48:6d:27 (oui Unknown), length 398
09:18:39.028519 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP, Request from 52:54:00:48:6d:27 (oui Unknown), length 398

I also tried to restart neutron-dhcp (q-dhcp) but no change. (If I manually configure a dhcpd (isc-dhcp-server) server this server responds to the dhcp boot requests).

I solved the problem. Or lets say at least the stack.sh script run without any error having the same settings like described in the manuals. I used ubuntu 14.04 server instead of 12.04...

"URI qemu:///system does not support events:" this message could be useful. Is libvirtd running? ps -aux |grep libvirt . Try a look into /var/log/libvirt/libvirtd.log

what's the output of nova-manage service list ? Are all services up and running?

Hi,

I'm trying ironic on devstack. My localrc configuration is the same like described in Deploying Ironic with devstack

Unfortunately I get the errors:

error: Failed to define network from /dev/fd/63
error: XML error: unexpected virtualport type -1

Output of stack.sh before the errors occur:

++ create_bridge_and_vms
++ sudo su ubuntu -c /home/ubuntu/devstack/tools/ironic/scripts/setup-network
+ LIBVIRT_CONNECT_URI=qemu:///system
+++ dirname /home/ubuntu/devstack/tools/ironic/scripts/setup-network
++ cd /home/ubuntu/devstack/tools/ironic/scripts/..
++ pwd
+ TOP_DIR=/home/ubuntu/devstack/tools/ironic
+ BRIDGE_SUFFIX=
+ BRIDGE_NAME=brbm
+ export VIRSH_DEFAULT_CONNECT_URI=qemu:///system
+ VIRSH_DEFAULT_CONNECT_URI=qemu:///system
+ sudo ovs-vsctl list-br
+ grep 'brbm$'
brbm
+ virsh net-list
+ grep 'brbm '
+ virsh net-list --inactive
+ grep 'brbm '
+ virsh net-define /dev/fd/63
++ sed s/brbm/brbm/ /home/ubuntu/devstack/tools/ironic/templates/brbm.xml

The brbm.xml file seems to be okay. The output of virsh net-list --inactive is

Name                 State      Autostart
-----------------------------------------
default              inactive   yes

And if I call the command virsh-netdefine /dev/fd/63 manually it fails due to "No such file or directory". Any ideas? Seems to be somthing with network configuration? Do I need special network settings for ironic?

Thanks Bill! I went one step further: I set up my devstack VM again, cloned devstack, run stack.sh an voila. It worked :-) but actually this isn't quite the solution of the problem I think.

It seems there is no network configured:

WARNING nova.network.quantumv2.api [req-66d78e14-0764-4f9d-b591-b9183761926d 99ae1e5b2ac34df1a9bd8d0522c7b409 39c9175513ec4bd8871182d6892f0164] [instance: 53ab53e0-f32c-4c5a-8bb3-9ef2a272f45b] No network configured!

Since you are using quantum you have to create a quantum net (quantum net-create) and also a subnet. quantum net-create vmnet quantum subnet-create vmnet 10.0.0.0/24

Moreover it seems your rabbitmq server isn't configured correctly.

 AMQP server on 172.16.77.130:5672 is unreachable:

Have you set up the rabbitmq according to http://docs.openstack.org/havana/install-guide/install/apt/content/basics-queue.html?

Hello,

I'm trying to intall devstack but the .stack.sh script always fails at the same point when calling the command glance image-create. I receive a HTTP500 Internal Server Error. Here are the last few lines of the output of stack.sh

This are the last few lines of screen-g-api.log:

I thought if mysql has something to do with it? I already increased the allocatable resources (buffer,cache) for mysql (/etc/mysql/my.cnf). Because I want to work with ironic due to localrc nova-network is disabled and neutron is enabled.

I already successfully set up devstack a few weeks ago, but since I installed my VM again I can't bring it up again.

I am a newbie to openstack. I am trying to install Openstack using devstack in my virtualbox (Ubuntu 12.04 32 bit). I ran ./stack.sh and get an httpinternalservererror. Please find the logs for reference here

--os-image-url http://10.0.2.15:9292 image-create --name cirros-0.3.1-x86_64-uec --is-public True --container-format ami --disk-format ami
2014-02-06 08:57:54 HTTPInternalServerError (HTTP 500)
2014-02-06 08:57:55 ++ failed
2014-02-06 08:57:55 ++ local r=1
2014-02-06 08:57:55 +++ jobs -p
2014-02-06 08:57:55 ++ kill
2014-02-06 08:57:55 ++ set +o xtrace
2014-02-06 08:57:55 stack.sh failed: full log in /opt/stack/logs/stack.sh.log.2014-02-06-084526

I just ran ./stack.sh and ended up in that error. I think this is an issue with glance. However, I do not know much about openstack, so i am unable to give more details. I have pasted the logs. Any help would be appreciated.

yes resetting might help - in fact it worked with me. BUT only when resetting into an active state: nova reset-state --active instancename

Hello,

I want to display hardware information, such as CPU temp, fan speed, IPMI infos, physical disk size etc. within openstack. Is there a suitable component for that? So far I looked into ironic and triple O but I'm in doubt if these components are the right ones for displaying hardware information.