On Fri, 20 Aug 2021 at 15:08, Dmitry Tantsur <dtantsur@redhat.com> wrote:

Hi,

On Fri, Aug 13, 2021 at 8:56 AM Anirudh Gupta <anyrude10@gmail.com> wrote:

Hi All,

I had a 900 GB hard disk on my Baremetal Node and it took approx 15 hours to make the baremetal node come in available state from clean_wait state.

Once the baremetal node came available, I was able to create a server and provision it with a user image.

Is taking 15 hours to erase_device in clean_wait normal for a 900 GB hard disk in Ironic?

Unfortunately, yes. If your hardware does not support ATA secure erase, the only way we can remove data is to shred the disk (essentially, write all 900 GB several times).

If you don't care about residual data on the disks, you can switch to metadata cleaning (only partition tables). This is fast but insecure. I don't know how the options are called in Kolla, but in Ironic you do something like this:

[deploy]
erase_devices_priority = 0
erase_devices_metadata_priority = 10

In kolla we do not try to own all options - simply create a config override file at /etc/kolla/config/ironic.conf including the above.

https://docs.openstack.org/kolla-ansible/latest/admin/advanced-configuration.html#openstack-service-configuration-in-kolla

Dmitry

 

Regards

Anirudh Gupta

On Mon, Aug 9, 2021 at 2:01 PM Anirudh Gupta <anyrude10@gmail.com> wrote:

Hi Mark,

Earlier I was passing the boot_mode as uefi while creating the baremetal node. 

On Kolla-Ansible Launchpad, I found some issues related to UEFI mode, so I didn't pass the parameter.

With IPXE and without passing UEFI boot mode parameter, my node started cleaning. It connected with the TFTP server.

But from the last 2 hours, the state is still in clean_wait only.

The ramdisk and kernel images I used were the ones mentioned in the link below

• https://tarballs.openstack.org/ironic-python-agent/dib/files/ipa-centos8-master.kernel
• https://tarballs.openstack.org/ironic-python-agent/dib/files/ipa-centos8-master.initramfs

For this I followed the latest kolla ansible document:-

• https://docs.openstack.org/kolla-ansible/latest/reference/bare-metal/ironic-guide.html

All I can see in ironic-conductor logs is:

2021-08-09 13:49:51.159 7 DEBUG ironic.drivers.modules.agent_base [-] Heartbeat from node 8b1ec553-fbc9-4912-bd33-88afc41b8f81 heartbeat /var/lib/kolla/venv/lib/python3.6/site-packages/ironic/drivers/modules/agent_base.py:641
2021-08-09 13:49:51.178 7 DEBUG ironic.drivers.modules.agent_client [-] Fetching status of agent commands for node 8b1ec553-fbc9-4912-bd33-88afc41b8f81 get_commands_status /var/lib/kolla/venv/lib/python3.6/site-packages/ironic/drivers/modules/agent_client.py:310
2021-08-09 13:49:51.186 7 DEBUG ironic.drivers.modules.agent_client [-] Status of agent commands for node 8b1ec553-fbc9-4912-bd33-88afc41b8f81: get_clean_steps: result "{'clean_steps': {'GenericHardwareManager': [{'step': 'erase_devices', 'priority': 10, 'interface': 'deploy', 'reboot_requested': False, 'abortable': True}, {'step': 'erase_devices_metadata', 'priority': 99, 'interface': 'deploy', 'reboot_requested': False, 'abortable': True}, {'step': 'erase_pstore', 'priority': 0, 'interface': 'deploy', 'reboot_requested': False, 'abortable': True}, {'step': 'delete_configuration', 'priority': 0, 'interface': 'raid', 'reboot_requested': False, 'abortable': True}, {'step': 'create_configuration', 'priority': 0, 'interface': 'raid', 'reboot_requested': False, 'abortable': True}, {'step': 'burnin_cpu', 'priority': 0, 'interface': 'deploy', 'reboot_requested': False, 'abortable': True}, {'step': 'burnin_disk', 'priority': 0, 'interface': 'deploy', 'reboot_requested': False, 'abortable': True}, {'step': 'burnin_memory', 'priority': 0, 'interface': 'deploy', 'reboot_requested': False, 'abortable': True}, {'step': 'burnin_network', 'priority': 0, 'interface': 'deploy', 'reboot_requested': False, 'abortable': True}]}, 'hardware_manager_version': {'MellanoxDeviceHardwareManager': '1', 'generic_hardware_manager': '1.1'}}", error "None"; execute_clean_step: result "{'clean_result': None, 'clean_step': {'step': 'erase_devices_metadata', 'priority': 99, 'interface': 'deploy', 'reboot_requested': False, 'abortable': True, 'requires_ramdisk': True}}", error "None"; execute_clean_step: result "None", error "None" get_commands_status /var/lib/kolla/venv/lib/python3.6/site-packages/ironic/drivers/modules/agent_client.py:342
2021-08-09 13:49:51.186 7 DEBUG ironic.drivers.modules.agent_base [-] Clean step still running for node 8b1ec553-fbc9-4912-bd33-88afc41b8f81: None _get_completed_command /var/lib/kolla/venv/lib/python3.6/site-packages/ironic/drivers/modules/agent_base.py:267

It would be a great help if you could suggest some pointers.

Regards

Anirudh Gupta

I tried 

On Mon, Aug 9, 2021 at 1:43 PM Mark Goddard <mark@stackhpc.com> wrote:

On Fri, 6 Aug 2021 at 13:49, Anirudh Gupta <anyrude10@gmail.com> wrote:

Hi Dmitry,

I tried taking TCPDUMP while the Baremetal Node was booting up and looked for tftp protocols and found there was some "File Not Found" traces for bootx64.efi

Then, I found a related post on openstack Discuss which suggested to enable IPXE

http://lists.openstack.org/pipermail/openstack-discuss/2019-October/010329.html

After re-deploying the setup with IPXE enabled, i found similar traces now for ipxe.efi file

Can you please now suggest what possibly could be a miss in configuration and steps to resolve it.

Hi Anirudh,

I'd suggest installing a tftp client on your machine and making some requests. The TFTP daemon runs in the ironic_pxe container, and TFTP files are served from /tftpboot in that container.

Mark

For your reference, I am attaching the complete tcpdump logs of both the Scenarios

Looking forward to hearing from you.

Regards

Anirudh Gupta

On Thu, Aug 5, 2021 at 4:56 PM Anirudh Gupta <anyrude10@gmail.com> wrote:

Hi Team,

On further debugging, I found an error in neutron-server logs

Failed to bind port 476d8175-ffc2-49ba-bb12-0a77c1f07e5f on host f4a43fa5-9c41-488e-a34d-714ae5a9d300 for vnic_type baremetal using segments [{'id': '1a5bbe96-2488-4971-925f-7c9346ba3ef5', 'network_type': 'flat', 'physical_network': 'physnet1', 'segmentation_id': None, 'network_id': '5b6cccec-ad86-4ed9-8d3c-72a31ec3a0d4'}]
2021-08-05 16:33:06.979 23 INFO neutron.plugins.ml2.plugin [req-54d11d51-7319-43ea-b70c-fe39d8aafe8a 21d6a238438e4294912746bcdc895e31 3eca725754e1405eb178cc39bd0da3aa - default default] Attempt 9 to bind port 476d8175-ffc2-49ba-bb12-0a77c1f07e5f

where 476d8175-ffc2-49ba-bb12-0a77c1f07e5f is the uuid of Baremetal Node

However the port is created in openstack, but its state is down

[ansible@localhost ~]$ openstack port list
+--------------------------------------+------+-------------------+---------------------------------------------------------------------------+--------+
| ID                                   | Name | MAC Address       | Fixed IP Addresses                                                        | Status |
+--------------------------------------+------+-------------------+---------------------------------------------------------------------------+--------+
| 07d6b83d-d83c-498f-8ba8-b4f21bef7249 |      | fa:16:3e:38:05:9d | ip_address='10.0.1.200', subnet_id='7b72c158-2146-4bd6-893b-bd76b4a3e869' | ACTIVE |
| 476d8175-ffc2-49ba-bb12-0a77c1f07e5f |      | 98:f2:b3:3f:72:d8 | ip_address='10.0.1.202', subnet_id='7b72c158-2146-4bd6-893b-bd76b4a3e869' | DOWN   |
+--------------------------------------+------+-------------------+---------------------------------------------------------------------------+--------+

98:f2:b3:3f:72:d8 is the mac address of my Baremetal Node on which PXE is enabled.

Can someone please help in resolving this issue.

Issue:

Node goes in clean_failed from clean_wait. 

Regards

Anirudh Gupta

On Tue, Aug 3, 2021 at 8:32 PM Anirudh Gupta <anyrude10@gmail.com> wrote:

Hi Dmitry,

I might be wrong, but as per my understanding if there would be an issue in dnsmasq, then IP 20.20.20.10 would not have been assigned to the machine.

TCPDUMP logs are as below:

20:16:58.938089 IP controller.bootps > 255.255.255.255.bootpc: BOOTP/DHCP, Reply, length 312
20:17:02.765291 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP, Request from 98:f2:b3:3f:72:e5 (oui Unknown), length 359
20:17:02.766303 IP controller.bootps > 255.255.255.255.bootpc: BOOTP/DHCP, Reply, length 312
20:17:26.944378 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP, Request from 98:f2:b3:3f:72:e5 (oui Unknown), length 347
20:17:26.944756 IP controller.bootps > 255.255.255.255.bootpc: BOOTP/DHCP, Reply, length 312
20:17:30.763627 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP, Request from 98:f2:b3:3f:72:e5 (oui Unknown), length 359
20:17:30.764620 IP controller.bootps > 255.255.255.255.bootpc: BOOTP/DHCP, Reply, length 312
20:17:54.938791 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP, Request from 98:f2:b3:3f:72:e5 (oui Unknown), length 347

Also the neutron dnsmasq logs and ironic inspector logs are attached in the mail.

Regards

Anirudh Gupta

On Tue, Aug 3, 2021 at 7:29 PM Dmitry Tantsur <dtantsur@redhat.com> wrote:

Hi,

You need to check the dnsmasq logs (there are two dnsmasqs: from neutron and from ironic-inspector). tcpdump may also help to determine where the packages are lost.

Dmitry

On Fri, Jul 30, 2021 at 10:29 PM Anirudh Gupta <anyrude10@gmail.com> wrote:

Hi Dmitry

Thanks for your time.

My system is getting IP 20.20.20.10 which is in the range defined in ironic_dnsmasq_dhcp_range field under globals.yml file.

ironic_dnsmasq_dhcp_range: "20.20.20.10,20.20.20.100"

And in the cleaning network (public1), the range defined is 20.20.20.150-20.20.20.200

As per my understanding, these 2 ranges should be mutually exclusive.

Please suggest if my understanding is not correct.

Any suggestions what should I do to resolve this issue?

Regards

Anirudh Gupta

On Sat, 31 Jul, 2021, 12:06 am Dmitry Tantsur, <dtantsur@redhat.com> wrote:

On Thu, Jul 29, 2021 at 6:05 PM Anirudh Gupta <anyrude10@gmail.com> wrote:

Hi Team,

In  to the email below, I have some updated information:-

Earlier the allocation range mentioned in "ironic_dnsmasq_dhcp_range" in globals.yml had an overlapping range with the cleaning network, due to which there was some issue in receiving the DHCP request 

After creating a cleaning network with a separate allocation range, I am successfully getting IP allocated to my Baremetal Node

• openstack subnet create subnet1 --network public1 --subnet-range 20.20.20.0/24 --allocation-pool start=20.20.20.150,end=20.20.20.200 --ip-version=4  --gateway=20.20.20.1 --dhcp

After getting the IP, there is no further action on the node. From "clean_wait", it goes into "clean_failed" state after around half an hour.

The IP address is not from the cleaning range, it may come from inspection. You probably need to investigate your network topology, maybe use tcpdump.

Unfortunately, I'm not fluent in Kolla to say if it can be a bug or not.

Dmitry

 

On verifying the logs, I could see the below error messages

• In /var/log/kolla/ironic/ironic-conductor.log, we observed the following error:

ERROR ironic.conductor.utils [-] Cleaning for node 3a56748e-a8ca-4dec-a332-ace18e6d494e failed. Timeout reached while cleaning the node. Please check if the ramdisk responsible for the cleaning is running on the node. Failed on step {}.

Note : For Cleaning the node, we have used the below images

 https://tarballs.openstack.org/ironic-python-agent/dib/files/ipa-centos8-master.kernel

 https://tarballs.openstack.org/ironic-python-agent/dib/files/ipa-centos8-master.initramfs

• In /var/log/kolla/nova/nova-compute-ironic.log, we observed the error

ERROR nova.compute.manager [req-810ffedf-3343-471c-94db-85411984e6cc - - - - -] No compute node record for host controller-ironic: nova.exception_Remote.ComputeHostNotFound_Remote: Compute host controller-ironic could not be found.

Can someone please help in this regard?

Regards

Anirudh Gupta

On Tue, Jul 27, 2021 at 12:52 PM Anirudh Gupta <anyrude10@gmail.com> wrote:

Hi Team,

We have deployed 2 node kolla ansible 12.0.0 in order to deploy openstack wallaby release. We have also enabled ironic in order to provision the bare metal nodes.

On each server we have 3 nics

• eno1 - OAM for external connectivity and endpoint's publicURL
• eno2 - Mgmt for internal communication between various openstack services.
• ens2f0 - Data Interface

Corresponding to this we have defined the following fields in globals.yml

• kolla_base_distro: "centos"
• kolla_install_type: "source"
• openstack_release: "wallaby"
• network_interface: "eno2"                               # MGMT interface
• kolla_external_vip_interface: "eno1"               # OAM Interface
• kolla_internal_vip_address: "192.168.10.3"    # MGMT Subnet free ip
• kolla_external_vip_address: "10.0.1.136"       # OAM subnet free IP
• neutron_external_interface: "ens2f0"             # Data Interface
• enable_neutron_provider_networks: "yes"

Note: Only relevant fields are being shown in this query

Also, for ironic following fields have been defined in globals.yml

• enable_ironic: "yes"
• enable_ironic_neutron_agent: "{{ enable_neutron | bool and enable_ironic | bool }}"
• enable_horizon_ironic: "{{ enable_ironic | bool }}"
• ironic_dnsmasq_interface: "ens2f0"                       # Data interface
• ironic_dnsmasq_dhcp_range: "20.20.20.10,20.20.20.100"
• ironic_dnsmasq_boot_file: "pxelinux.0"
• ironic_cleaning_network: "public1"
• ironic_dnsmasq_default_gateway: "20.20.20.1"

After successful deployment, a flat provider network with the name public1 is being created in openstack using the below commands:

• openstack network create public1 --provider-network-type flat --provider-physical-network physnet1
• openstack subnet create subnet1 --network public1 --subnet-range 20.20.20.0/24 --allocation-pool start=20.20.20.10,end=20.20.20.100 --ip-version=4  --gateway=20.20.20.1 --dhcp

Issue/Queries:

• Is the configuration done in globals.yml correct or is there anything else that needs to be done in order to separate control and data plane traffic?

• Also I have set automated_cleaning as "true" in ironic-conductor conatiner settings.But after creating the baremetal node, we run "node manage" command which runs successfully. Running "openstack baremetal node provide <node id>" command powers on the machine, sets the boot mode on Network Boot but no DHCP request for that particular mac is obtained on the controller. Is there anything I am missing that needs to be done in order to make ironic work?

Note: I have also verified that the nic is PXE enabled in system configuration setting

Regards

Anirudh Gupta

--

Red Hat GmbH, https://de.redhat.com/ , Registered seat: Grasbrunn,
Commercial register: Amtsgericht Muenchen, HRB 153243,
Managing Directors: Charles Cachera, Brian Klemm, Laurie Krebs, Michael O'Neill

--

Red Hat GmbH, https://de.redhat.com/ , Registered seat: Grasbrunn,
Commercial register: Amtsgericht Muenchen, HRB 153243,
Managing Directors: Charles Cachera, Brian Klemm, Laurie Krebs, Michael O'Neill

--

Red Hat GmbH, https://de.redhat.com/ , Registered seat: Grasbrunn,
Commercial register: Amtsgericht Muenchen, HRB 153243,
Managing Directors: Charles Cachera, Brian Klemm, Laurie Krebs, Michael O'Neill