Patrick MAMIA's profile - activity

To retrive ur uuid you can doxe sr-list and copy the uuid of your nfs iso storage. Now i'm facing another issues to launch the instances 'XENAPI_PLUGIN_FAILURE', 'network_config', 'PluginError', 'add bridge failed: Package not installed If someone can help ... and about that cinder ascii bug.thk

Hello for now i do not fix the issues about attaching volume. For the second one i find that even if got a SR NFS ISO on xenserver i need to update one setting to other-config:i18n-key=local-storage-iso and you can do that with xe sr-param-set uuid=*UUID* other-config:i18n-key=local-storage-iso

Hello My env is

1. 3 nodes (controller, network, compute)
2. the hypervisor is xenserver
3. cinder is installed on controller node
4. my cinder my configuration for backend is [lvm] volume_backend_name=lvm volume_driver=cinder.volume.drivers.lvm.LVMVolumeDriver iscsi_ip_address=controller iscsi_protocol=iscsi iscsi_helper=lioadm volume_group=cinder-volumes volumes_dir=/var/lib/cinder/volumes

i try to lauchn instances and get error at block device mapping step. When i look at the volume log of cinder i get

sudo cinder-rootwrap /etc/cinder/rootwrap.conf cinder-rtstool create /dev/cinder-volumes/volume-1fd8e239-7cf7-4607-9213-ca7d1c5831b3 iqn.2010-10.org.openstack:volume-1fd8e239-7cf7-4607-9213-ca7d1c5831b3 fNuAoqyQGh6dFiBYRKTn aSRMHhkRUDzds5iG False -p3260 -acontroller

> ERROR cinder.volume.targets.lio Stderr: u'Traceback (most recent call last):\n File "/bin/cinder-rtstool", line 10, in <module>\n sys.exit(main())\n File "/usr/lib/python2.7/site-packages/cinder/cmd/rtstool.py", line 287, in main\n **optional_args)\n File "/usr/lib/python2.7/site-packages/cinder/cmd/rtstool.py", line 92, in create\n % {\'msg_type\': msg_type, \'port\': portals_port, \'ip\': ip})\nUnicodeEncodeError: \'ascii\' codec can\'t encode character u\'\\xe9\' in position 10: ordinal not in range(128)\n'

When i try to launch an instance qithout creating a volume i got this error on nova-compute.log

Cannot find SR of content-type ISO

But i do have a NFS ISO on my xenserver hypervisor

Finally i don't know which disk have to get space (xenserver ou compute or controller). my cinder-volume is the controller for now and i have two computes nodes. @brucema or @Bernd Bausch any help? since xenserver has more than 300gb free, cinder volume 99gb and / 45gb but hypervisor stat show these

| local_gb | 1016 | | local_gb_used | 0 | | memory_mb | 65444 | | memory_mb_used | 1024 | | running_vms | 0 | | vcpus | 80 | | vcpus_used | 0

openstack hypervisor stats show +----------------------+-------+ | Field | Value | +----------------------+-------+ | count | 2 | | current_workload | 0 | | disk_available_least | -3544 | | free_disk_gb | 1016 | | free_ram_mb | 64420 |

DEBUG nova.scheduler.filters.disk_filter - default default] (compute, xenserver2) ram: 32210MB disk: -1814528MB io_ops: 0 instances: 0 does not have 1024 MB usable disk, it only has -1814528.0 MB usable disk. host_passes

Hello,

I faced the same issue, if it can help, take a look here

https://ask.openstack.org/en/question/121754/cannot-see-new-compute-node-in-nova-service-list/?answer=121885#post-id-121885 (Cannot see new compute node in nova service-list - Resolved)

finally i manually create the vxlan port on the tunnel bridgewith this command

 ovs-vsctl add-port br-tun vxlan-c0a80217 -- set interface vxlan-c0a80217 type=vxlan     options:remote_ip=[ip_of_network_node] options:local_ip=[ip_of_this_node] options:in_key=flow options:out_key=flow options:df_default=true

Hello, I'm happy to tell you i partially find a solution.

First of all ensure yourself that in the /etc/nova/nova.conf of the new compute node you have

enabled_apis = osapi_compute,metadata

After this, you have to make sur that you modify all variable of nova server on your new compute node to point to controller node that have the API, OS-COMPUTE, METADATA,.. example for me i changed these values on the nova.conf of the new compute node

osapi_compute_listen=controller
xvpvncproxy_host=controller
api_servers=http://controller:9292

Make sure that you have a correct transport_url value and connection values for [placement_database], [database] and [api_database] with correct nova databases nova_placement, nova, nova_api.

After all of this i execute the command on the controller node

nova-manage cell_v2 discover_hosts  --cell_uuid [uuid_of_my_cell] --verbose

and the new compute node appears on the openstack compute service list.

i get the uuid of my cell with the command below

 nova-manage cell_v2 list_cells --verbose

Only one bug still remain on nova-computelog on the new compute node

ResourceProviderCreationFailed: Failed to create resource provider xenserver2

i imagine that's because my two compute nodes has the same resource provider for the moment and the first node compute already register xenserver2 as his hypervisor hostname.

@Bernd Bausch any help?

hey @brucema i'm trying to build another compute since, i got some issues to make it discover by the controller node so when i success i try your suggestions. Thanks

that is what i got as result when i try nova-manage cell_v2 discover_hosts at controller Skipping cell0 since it does not contain hosts. Getting computes from cell 'default': xx Checking host mapping for compute host 'compute': xx Found 0 unmapped computes in

I'm trying to add 2nd compute

Also in nova-api-placement.log in controller node it seem to have communication between compute2 and controller because i saw 4 lines with the ip of compute2 nodes with status 200 so it seem that there is communication. nova-status upgrade check show me Cells V2: Success

@Bernd Bausch what i mean is that nova-compute service is active on compute2 node and but this service do not appear on nova service list at controller node. Nova-compute.log on compute node don not show any error even in debug mode. so finally these compute2 host do not appear in openstack

@Bernd Bausch in nova.api.openstack.placement on controller node, i saw 4 lines with the ip of compute2 nodes it seem that there is communication. There is some endpoint of /placement/resource_providers/ like /allocations, /inventories, /aggregates, /traits with status 200 no error on nova-compute

Hello,

I setup a 3 nodes openstack installation (compute +controller + network). Everything seem to be ok with neutron agent-list and nova service-list. i create another node with hostname compute2 and setup evrything, it seem to be ok on this node because nova-compute is running, openvswitch-agent on this node is up and alive in the controller node but when i make su -s /bin/sh -c "nova-manage cell_v2 discover_hosts --verbose" nova it only show me this

Skipping cell0 since it does not contain hosts. Getting computes from cell 'default': 879d48ff-e984-43df-ac92-ffdf4e08fa01 Checking host mapping for compute host 'compute': 52061cc8-bd27-4659-aba4-be81eb5d8a0e Found 0 unmapped computes in cell: 879d48ff-e984-43df-ac92-ffdf4e08fa01

the compute2 is not listed in service list. I take a look in rabbitmq loq and i see the request from that node (it's ip) is received and accept. i look in other log file in controller node and i see the ip of that node like there is communication between the new compute2 node and the controller node but it still don't appear on nova service-list.

Could someone explain me what's wrong. I admit that i begin with openstack and i dont really understand how everything work, like these stuff of nova cell. Could you help me please?

Thanks

Hello Do someone find a solution for that issues? I got the same problem since few days and try many solutions without success. I got 4 Nodes (controller + network + 2 computes).

@brucema ..your speculation is right i use packstack and use the default Centos partitioning scheme, u seem to have experience on that issue, some advices to avoid it?

@Bernd Bausch hey, where can i enable that debug logging, which files and which lines? thanks

If i supposed that my config in neutron.conf, ml2.ini and openvswitch-agent.ini are correct, which command should build that correct config for tunneling.

Hello @Bernd Bausch

Thk for ur response. What i mean is that after my first installation with packstack the br-tun with local_ip and remote_ip was there but now there is a br-tun brigde without that specific port. I already look at openvswitch-agent.log, there are some lines as br-tun has datapath

I setup a packstack installation between three nodes (compute, network and controller). I need to make some customs setting and after it i have to restart neutron agent locate at network and compute. After some debug i could see all my agent with command

neutron agent-list

There all up and enable but when i look at ovs-vsctl show (on compute or network) there br-tun that was there with local_ip and remote_ip. i thinked that restart openvswitch agent will resolve it but it doesn't