Install conditionalizations

OVS related deployment conditionalizations.
Patchset 1 review updates.
Updates based on additional inputs.
Patchset 3 review updates.
Fixed some unexpanded substitutions and formatting issues throughout.
Patchset 5 updates.

Signed-off-by: Ron Stone <ronald.stone@windriver.com>
Change-Id: Ib86bf0e13236a40f7a472d4448a9b2d3cc165deb
Signed-off-by: Ron Stone <ronald.stone@windriver.com>

Reorg OpenStack installion for DS consumption

This review replaces https://review.opendev.org/c/starlingx/docs/+/801130

Signed-off-by: Ron Stone <ronald.stone@windriver.com>
Change-Id: Iab9c8d56cff9c1bc57e7e09fb3ceef7cf626edad
Signed-off-by: Ron Stone <ronald.stone@windriver.com>
This commit is contained in:
Ron Stone 2021-07-16 15:37:55 -04:00
parent f6180faca7
commit d8d90b4d75
27 changed files with 1153 additions and 539 deletions

View File

@ -1,5 +1,6 @@
{
"restructuredtext.languageServer.disabled": true,
"restructuredtext.preview.sphinx.disabled": true,
"restructuredtext.linter.disabled": true
"restructuredtext.linter.disabled": true,
"restructuredtext.confPath": ""
}

View File

@ -70,3 +70,6 @@
..
.. |installer-image-name| replace:: bootimage
.. |OVS-DPDK| replace:: |OVS|-|DPDK|
.. |ovs-dpdk| replace:: ovs-dpdk

View File

@ -127,9 +127,15 @@ Configure worker nodes
done
When using |OVS|-|DPDK|, configure 1x 1G huge page for vSwitch memory on
each |NUMA| node where vswitch is running on this host, with the
following command:
When using |OVS|-|DPDK|, configure 1G of huge pages for vSwitch memory on
each |NUMA| node where vswitch is running on the host. It is recommended
to configure 1x 1G huge page (-1G 1) for vSwitch memory on each |NUMA|
node where vswitch is running on host.
However, due to a limitation with Kubernetes, only a single huge page
size is supported on any one host. If your application VMs require 2M
huge pages, then configure 500x 2M huge pages (-2M 500) for vSwitch
memory on each |NUMA| node where vswitch is running on host.
.. code-block:: bash
@ -150,8 +156,9 @@ Configure worker nodes
huge pages to enable networking and must use a flavor with property:
hw:mem_page_size=large
Configure the huge pages for |VMs| in an |OVS|-|DPDK| environment for
this host with the command:
Configure the huge pages for |VMs| in an |OVS-DPDK| environment on
this host, assuming 1G huge page size is being used on this host, with
the following commands:
.. code-block:: bash
@ -183,7 +190,14 @@ Configure worker nodes
# system host-show ${NODE} --nowrap | fgrep rootfs )
# Create new PARTITION on selected disk, and take note of new partitions uuid in response
PARTITION_SIZE=34 # Use default of 34G for this nova-local partition
# The size of the PARTITION needs to be large enough to hold the aggregate size of
# all nova ephemeral disks of all VMs that you want to be able to host on this host,
# but is limited by the size and space available on the physical disk you chose above.
# The following example uses a small PARTITION size such that you can fit it on the
# root disk, if that is what you chose above.
# Additional PARTITION(s) from additional disks can be added later if required.
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol ${NODE} <disk-uuid> ${PARTITION_SIZE}
# Add new partition to nova-local local volume group
@ -272,7 +286,8 @@ Optionally Configure PCI-SRIOV Interfaces
system host-if-modify -m 1500 -n sriov0 -c pci-sriov ${NODE} <sriov0-if-uuid>
system host-if-modify -m 1500 -n sriov1 -c pci-sriov ${NODE} <sriov1-if-uuid>
# Create Data Networks that the 'pci-sriov' interfaces will be connected to
# If not already created, create Data Networks that the 'pci-sriov'
# interfaces will be connected to
DATANET0='datanet0'
DATANET1='datanet1'
system datanetwork-add ${DATANET0} vlan
@ -283,8 +298,8 @@ Optionally Configure PCI-SRIOV Interfaces
system interface-datanetwork-assign ${NODE} <sriov1-if-uuid> ${DATANET1}
* To enable using |SRIOV| network attachments for the above interfaces in
Kubernetes hosted application containers:
* **For Kubernetes only** To enable using |SRIOV| network attachments for
the above interfaces in Kubernetes hosted application containers:
* Configure the Kubernetes |SRIOV| device plugin.

View File

@ -39,7 +39,8 @@ Bootstrap system on controller-0
--------------------------------
#. Login using the username / password of "sysadmin" / "sysadmin".
When logging in for the first time, you will be forced to change the password.
When logging in for the first time, you will be forced to change the
password.
::
@ -107,9 +108,9 @@ Bootstrap system on controller-0
#. Create a minimal user configuration override file.
To use this method, create your override file at ``$HOME/localhost.yml``
and provide the minimum required parameters for the deployment configuration
as shown in the example below. Use the OAM IP SUBNET and IP ADDRESSing
applicable to your deployment environment.
and provide the minimum required parameters for the deployment
configuration as shown in the example below. Use the OAM IP SUBNET and IP
ADDRESSing applicable to your deployment environment.
::
@ -148,7 +149,7 @@ Bootstrap system on controller-0
:start-after: docker-reg-begin
:end-before: docker-reg-end
.. code-block::
.. code-block:: yaml
docker_registries:
quay.io:
@ -187,7 +188,7 @@ Bootstrap system on controller-0
:start-after: firewall-begin
:end-before: firewall-end
.. code-block::
.. code-block:: bash
# Add these lines to configure Docker to use a proxy server
docker_http_proxy: http://my.proxy.com:1080
@ -222,44 +223,50 @@ Configure controller-0
#. Configure the |OAM| interface of controller-0 and specify the
attached network as "oam".
Use the |OAM| port name that is applicable to your deployment environment,
for example eth0:
The following example configures the |OAM| interface on a physical untagged
ethernet port. Use the |OAM| port name that is applicable to your
deployment environment, for example eth0:
::
.. code-block:: bash
OAM_IF=<OAM-PORT>
system host-if-modify controller-0 $OAM_IF -c platform
system interface-network-assign controller-0 $OAM_IF oam
To configure a vlan or aggregated ethernet interface, see :ref:`Node
Interfaces <node-interfaces-index>`.
#. Configure the MGMT interface of controller-0 and specify the attached
networks of both "mgmt" and "cluster-host".
Use the MGMT port name that is applicable to your deployment environment,
for example eth1:
The following example configures the MGMT interface on a physical untagged
ethernet port. Use the MGMT port name that is applicable to your deployment
environment, for example eth1:
.. code-block:: none
.. code-block:: bash
MGMT_IF=<MGMT-PORT>
# De-provision loopback interface and
# remove mgmt and cluster-host networks from loopback interface
system host-if-modify controller-0 lo -c none
IFNET_UUIDS=$(system interface-network-list controller-0 | awk '{if ($6=="lo") print $4;}')
for UUID in $IFNET_UUIDS; do
system interface-network-remove ${UUID}
done
# Configure management interface and assign mgmt and cluster-host networks to it
system host-if-modify controller-0 $MGMT_IF -c platform
system interface-network-assign controller-0 $MGMT_IF mgmt
system interface-network-assign controller-0 $MGMT_IF cluster-host
To configure a vlan or aggregated ethernet interface, see :ref:`Node
Interfaces <node-interfaces-index>`.
#. Configure |NTP| servers for network time synchronization:
::
system ntp-modify ntpservers=0.pool.ntp.org,1.pool.ntp.org
To configure |PTP| instead of |NTP|, see ref:`PTP Server Configuration
<ptp-server-config-index>`.
.. only:: openstack
*************************************
@ -281,71 +288,125 @@ Configure controller-0
system host-label-assign controller-0 openvswitch=enabled
system host-label-assign controller-0 sriov=enabled
#. **For OpenStack only:** Due to the additional openstack services running
on the |AIO| controller platform cores, a minimum of 4 platform cores are
required, 6 platform cores are recommended.
Increase the number of platform cores with the following commands:
.. code-block::
# assign 6 cores on processor/numa-node 0 on controller-0 to platform
system host-cpu-modify -f platform -p0 6 controller-0
#. Due to the additional openstack services containers running on the
controller host, the size of the docker filesystem needs to be
increased from the default size of 30G to 60G.
.. code-block:: bash
# check existing size of docker fs
system host-fs-list controller-1
# check available space (Avail Size (GiB)) in cgts-vg LVG where docker fs is located
system host-lvg-list controller-1
# if existing docker fs size + cgts-vg available space is less than 60G,
# you will need to add a new disk partition to cgts-vg
# Assuming you have unused space on ROOT DISK, add partition to ROOT DISK.
# ( if not use another unused disk )
# Get device path of ROOT DISK
system host-show controller-1 --nowrap | fgrep rootfs
# Get UUID of ROOT DISK by listing disks
system host-disk-list controller-1
# Create new PARTITION on ROOT DISK, and take note of new partitions uuid in response
# Use a partition size such that youll be able to increase docker fs size from 30G to 60G
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol controller-1 <root-disk-uuid> ${PARTITION_SIZE}
# Add new partition to cgts-vg local volume group
system host-pv-add controller-1 cgts-vg <NEW_PARTITION_UUID>
sleep 2 # wait for partition to be added
# Increase docker filesystem to 60G
system host-fs-modify controller-1 docker=60
#. **For OpenStack only:** Configure the system setting for the vSwitch.
.. only:: starlingx
StarlingX has |OVS| (kernel-based) vSwitch configured as default:
* Runs in a container; defined within the helm charts of stx-openstack
manifest.
* Shares the core(s) assigned to the platform.
If you require better performance, |OVS|-|DPDK| (|OVS| with the Data
If you require better performance, |OVS-DPDK| (|OVS| with the Data
Plane Development Kit, which is supported only on bare metal hardware)
should be used:
* Runs directly on the host (it is not containerized).
* Requires that at least 1 core be assigned/dedicated to the vSwitch function.
Requires that at least 1 core be assigned/dedicated to the vSwitch
function.
**To deploy the default containerized OVS:**
To deploy the default containerized |OVS|:
::
system modify --vswitch_type none
This does not run any vSwitch directly on the host, instead, it uses the
containerized |OVS| defined in the helm charts of stx-openstack
This does not run any vSwitch directly on the host, instead, it uses
the containerized |OVS| defined in the helm charts of stx-openstack
manifest.
**To deploy OVS-DPDK, run the following command:**
To deploy |OVS-DPDK|, run the following command:
::
.. parsed-literal::
system modify --vswitch_type ovs-dpdk
system modify --vswitch_type |ovs-dpdk|
Default recommendation for an |AIO|-controller is to use a single core
for |OVS|-|DPDK| vswitch.
::
Default recommendation for an |AIO|-controller is to use a single
core |OVS-DPDK| vswitch.
.. code-block:: bash
# assign 1 core on processor/numa-node 0 on controller-0 to vswitch
system host-cpu-modify -f vswitch -p0 1 controller-0
system host-cpu-modify -f vswitch -p0 0 controller-0
Once vswitch_type is set to |OVS|-|DPDK|, any subsequent nodes created
will default to automatically assigning 1 vSwitch core for |AIO|
controllers and 2 vSwitch cores for compute-labeled worker nodes.
Once vswitch_type is set to |OVS-DPDK|, any subsequent nodes created will
default to automatically assigning 1 vSwitch core for |AIO| controllers
and 2 vSwitch cores (1 on each numa-node) for compute-labeled worker
nodes.
When using |OVS-DPDK|, configure 1G huge page for vSwitch memory on each
|NUMA| node where vswitch is running on this host. It is recommended to
configure 1x 1G huge page (-1G 1) for vSwitch memory on each |NUMA| node
where vswitch is running on host.
However, due to a limitation with Kubernetes, only a single huge page
size is supported on any one host. If your application |VMs| require 2M
huge pages, then configure 500x 2M huge pages (-2M 500) for vSwitch
memory on each |NUMA| node where vswitch is running on host.
When using |OVS|-|DPDK|, configure 1x 1G huge page for vSwitch memory on
each |NUMA| node where vswitch is running on this host, with the
following command:
::
.. code-block::
# assign 1x 1G huge page on processor/numa-node 0 on controller-0 to vswitch
system host-memory-modify -f vswitch -1G 1 controller-0 0
.. important::
|VMs| created in an |OVS|-|DPDK| environment must be configured to use
|VMs| created in an |OVS-DPDK| environment must be configured to use
huge pages to enable networking and must use a flavor with property:
hw:mem_page_size=large
``hw:mem_page_size=large``
Configure the huge pages for |VMs| in an |OVS|-|DPDK| environment on this host with
the commands:
Configure the huge pages for |VMs| in an |OVS-DPDK| environment on
this host, assuming 1G huge page size is being used on this host, with
the following commands:
::
.. code-block:: bash
# assign 10x 1G huge page on processor/numa-node 0 on controller-0 to applications
system host-memory-modify -f application -1G 10 controller-0 0
@ -353,6 +414,7 @@ Configure controller-0
# assign 10x 1G huge page on processor/numa-node 1 on controller-0 to applications
system host-memory-modify -f application -1G 10 controller-0 1
.. note::
After controller-0 is unlocked, changing vswitch_type requires
@ -363,6 +425,7 @@ Configure controller-0
.. code-block:: bash
# Create nova-local local volume group
system host-lvg-add ${NODE} nova-local
# Get UUID of DISK to create PARTITION to be added to nova-local local volume group
@ -375,7 +438,13 @@ Configure controller-0
# system host-show ${NODE} --nowrap | fgrep rootfs )
# Create new PARTITION on selected disk, and take note of new partitions uuid in response
PARTITION_SIZE=34 # Use default of 34G for this nova-local partition
# The size of the PARTITION needs to be large enough to hold the aggregate size of
# all nova ephemeral disks of all VMs that you want to be able to host on this host,
# but is limited by the size and space available on the physical disk you chose above.
# The following example uses a small PARTITION size such that you can fit it on the
# root disk, if that is what you chose above.
# Additional PARTITION(s) from additional disks can be added later if required.
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol ${NODE} <disk-uuid> ${PARTITION_SIZE}
# Add new partition to nova-local local volume group
@ -389,11 +458,12 @@ Configure controller-0
.. important::
A compute-labeled All-in-one controller host **MUST** have at least one Data class interface.
A compute-labeled All-in-one controller host **MUST** have at least
one Data class interface.
* Configure the data interfaces for controller-0.
::
.. code-block:: bash
export NODE=controller-0
@ -439,7 +509,7 @@ Optionally Configure PCI-SRIOV Interfaces
* Configure the pci-sriov interfaces for controller-0.
::
.. code-block:: bash
export NODE=controller-0
@ -457,7 +527,8 @@ Optionally Configure PCI-SRIOV Interfaces
system host-if-modify -m 1500 -n sriov0 -c pci-sriov ${NODE} <sriov0-if-uuid>
system host-if-modify -m 1500 -n sriov1 -c pci-sriov ${NODE} <sriov1-if-uuid>
# Create Data Networks that the 'pci-sriov' interfaces will be connected to
# If not already created, create Data Networks that the 'pci-sriov'
# interfaces will be connected to
DATANET0='datanet0'
DATANET1='datanet1'
system datanetwork-add ${DATANET0} vlan
@ -468,8 +539,8 @@ Optionally Configure PCI-SRIOV Interfaces
system interface-datanetwork-assign ${NODE} <sriov1-if-uuid> ${DATANET1}
* To enable using |SRIOV| network attachments for the above interfaces in
Kubernetes hosted application containers:
* **For Kubernetes Only:** To enable using |SRIOV| network attachments for
the above interfaces in Kubernetes hosted application containers:
* Configure the Kubernetes |SRIOV| device plugin.
@ -481,7 +552,7 @@ Optionally Configure PCI-SRIOV Interfaces
containers on this host, configure the number of 1G Huge pages required
on both |NUMA| nodes.
::
.. code-block:: bash
# assign 10x 1G huge page on processor/numa-node 0 on controller-0 to applications
system host-memory-modify -f application controller-0 0 -1G 10
@ -611,8 +682,9 @@ Configure controller-1
#. Configure the |OAM| interface of controller-1 and specify the
attached network of "oam".
Use the |OAM| port name that is applicable to your deployment environment,
for example eth0:
The following example configures the |OAM| interface on a physical untagged
ethernet port, use the |OAM| port name that is applicable to your
deployment environment, for example eth0:
::
@ -620,6 +692,9 @@ Configure controller-1
system host-if-modify controller-1 $OAM_IF -c platform
system interface-network-assign controller-1 $OAM_IF oam
To configure a vlan or aggregated ethernet interface, see :ref:`Node
Interfaces <node-interfaces-index>`.
#. The MGMT interface is partially set up by the network install procedure;
configuring the port used for network install as the MGMT port and
specifying the attached network of "mgmt".
@ -639,11 +714,11 @@ Configure controller-1
.. important::
**These steps are required only if the StarlingX OpenStack application
(stx-openstack) will be installed.**
These steps are required only if the |prod-os| application
(|prefix|-openstack) will be installed.
#. **For OpenStack only:** Assign OpenStack host labels to controller-1 in
support of installing the stx-openstack manifest and helm-charts later.
support of installing the |prefix|-openstack manifest and helm-charts later.
::
@ -652,12 +727,57 @@ Configure controller-1
system host-label-assign controller-1 openvswitch=enabled
system host-label-assign controller-1 sriov=enabled
#. **For OpenStack only:** Due to the additional openstack services running
on the |AIO| controller platform cores, a minimum of 4 platform cores are
required, 6 platform cores are recommended.
Increase the number of platform cores with the following commands:
.. code-block::
# assign 6 cores on processor/numa-node 0 on controller-1 to platform
system host-cpu-modify -f platform -p0 6 controller-1
#. Due to the additional openstack services containers running on the
controller host, the size of the docker filesystem needs to be
increased from the default size of 30G to 60G.
.. code-block:: bash
# check existing size of docker fs
system host-fs-list controller-1
# check available space (Avail Size (GiB)) in cgts-vg LVG where docker fs is located
system host-lvg-list controller-1
# if existing docker fs size + cgts-vg available space is less than 60G,
# you will need to add a new disk partition to cgts-vg
# Assuming you have unused space on ROOT DISK, add partition to ROOT DISK.
# ( if not use another unused disk )
# Get device path of ROOT DISK
system host-show controller-1 --nowrap | fgrep rootfs
# Get UUID of ROOT DISK by listing disks
system host-disk-list controller-1
# Create new PARTITION on ROOT DISK, and take note of new partitions uuid in response
# Use a partition size such that youll be able to increase docker fs size from 30G to 60G
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol controller-1 <root-disk-uuid> ${PARTITION_SIZE}
# Add new partition to cgts-vg local volume group
system host-pv-add controller-1 cgts-vg <NEW_PARTITION_UUID>
sleep 2 # wait for partition to be added
# Increase docker filesystem to 60G
system host-fs-modify controller-1 docker=60
#. **For OpenStack only:** Configure the host settings for the vSwitch.
**If using OVS-DPDK vswitch, run the following commands:**
If using |OVS-DPDK| vswitch, run the following commands:
Default recommendation for an AIO-controller is to use a single core
for |OVS|-|DPDK| vswitch. This should have been automatically configured,
Default recommendation for an |AIO|-controller is to use a single core
for |OVS-DPDK| vswitch. This should have been automatically configured,
if not run the following command.
.. code-block:: bash
@ -666,9 +786,15 @@ Configure controller-1
system host-cpu-modify -f vswitch -p0 1 controller-1
When using |OVS|-|DPDK|, configure 1x 1G huge page for vSwitch memory on
each |NUMA| node where vswitch is running on this host, with the
following command:
When using |OVS-DPDK|, configure 1G of huge pages for vSwitch memory on
each |NUMA| node where vswitch is running on the host. It is recommended
to configure 1x 1G huge page (-1G 1) for vSwitch memory on each |NUMA|
node where vswitch is running on host.
However, due to a limitation with Kubernetes, only a single huge page
size is supported on any one host. If your application VMs require 2M
huge pages, then configure 500x 2M huge pages (-2M 500) for vSwitch
memory on each |NUMA| node where vswitch is running on host.
.. code-block:: bash
@ -678,12 +804,13 @@ Configure controller-1
.. important::
|VMs| created in an |OVS|-|DPDK| environment must be configured to use
|VMs| created in an |OVS-DPDK| environment must be configured to use
huge pages to enable networking and must use a flavor with property:
hw:mem_page_size=large
Configure the huge pages for |VMs| in an |OVS|-|DPDK| environment for
this host with the command:
Configure the huge pages for |VMs| in an |OVS-DPDK| environment on
this host, assuming 1G huge page size is being used on this host, with
the following commands:
.. code-block:: bash
@ -699,6 +826,7 @@ Configure controller-1
.. code-block:: bash
# Create nova-local local volume group
system host-lvg-add ${NODE} nova-local
# Get UUID of DISK to create PARTITION to be added to nova-local local volume group
@ -711,7 +839,14 @@ Configure controller-1
# system host-show ${NODE} --nowrap | fgrep rootfs )
# Create new PARTITION on selected disk, and take note of new partitions uuid in response
PARTITION_SIZE=34 # Use default of 34G for this nova-local partition
# The size of the PARTITION needs to be large enough to hold the aggregate size of
# all nova ephemeral disks of all VMs that you want to be able to host on this host,
# but is limited by the size and space available on the physical disk you chose above.
# The following example uses a small PARTITION size such that you can fit it on the
# root disk, if that is what you chose above.
# Additional PARTITION(s) from additional disks can be added later if required.
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol ${NODE} <disk-uuid> ${PARTITION_SIZE}
# Add new partition to nova-local local volume group
@ -793,7 +928,8 @@ Optionally Configure PCI-SRIOV Interfaces
system host-if-modify -m 1500 -n sriov0 -c pci-sriov ${NODE} <sriov0-if-uuid>
system host-if-modify -m 1500 -n sriov1 -c pci-sriov ${NODE} <sriov1-if-uuid>
# Create Data Networks that the 'pci-sriov' interfaces will be connected to
# If not already created, create Data Networks that the 'pci-sriov' interfaces
# will be connected to
DATANET0='datanet0'
DATANET1='datanet1'
system datanetwork-add ${DATANET0} vlan
@ -804,8 +940,8 @@ Optionally Configure PCI-SRIOV Interfaces
system interface-datanetwork-assign ${NODE} <sriov1-if-uuid> ${DATANET1}
* To enable using |SRIOV| network attachments for the above interfaces in
Kubernetes hosted application containers:
* **For Kubernetes only:** To enable using |SRIOV| network attachments for
the above interfaces in Kubernetes hosted application containers:
* Configure the Kubernetes |SRIOV| device plugin.

View File

@ -222,8 +222,9 @@ The newly installed controller needs to be configured.
source /etc/platform/openrc
#. Configure the |OAM| interface of controller-0 and specify the attached
network as "oam". Use the |OAM| port name that is applicable to your
deployment environment, for example eth0:
network as "oam". The following example configures the OAM interface on a
physical untagged ethernet port, use |OAM| port name that is applicable to
your deployment environment, for example eth0:
::
@ -231,12 +232,17 @@ The newly installed controller needs to be configured.
system host-if-modify controller-0 $OAM_IF -c platform
system interface-network-assign controller-0 $OAM_IF oam
To configure a vlan or aggregated ethernet interface, see :ref:`Node
Interfaces <node-interfaces-index>`.
#. Configure |NTP| servers for network time synchronization:
::
system ntp-modify ntpservers=0.pool.ntp.org,1.pool.ntp.org
To configure |PTP| instead of |NTP|, see :ref:`PTP Server Configuration
<ptp-server-config-index>`.
.. only:: openstack
*************************************
@ -260,62 +266,120 @@ The newly installed controller needs to be configured.
system host-label-assign controller-0 openvswitch=enabled
system host-label-assign controller-0 sriov=enabled
#. **For OpenStack only:** Due to the additional openstack services running
on the |AIO| controller platform cores, a minimum of 4 platform cores are
required, 6 platform cores are recommended.
Increase the number of platform cores with the following commands:
.. code-block::
# Assign 6 cores on processor/numa-node 0 on controller-0 to platform
system host-cpu-modify -f platform -p0 6 controller-0
#. Due to the additional openstack services containers running on the
controller host, the size of the docker filesystem needs to be
increased from the default size of 30G to 60G.
.. code-block:: bash
# check existing size of docker fs
system host-fs-list controller-1
# check available space (Avail Size (GiB)) in cgts-vg LVG where docker fs is located
system host-lvg-list controller-1
# if existing docker fs size + cgts-vg available space is less than 60G,
# you will need to add a new disk partition to cgts-vg
# Assuming you have unused space on ROOT DISK, add partition to ROOT DISK.
# ( if not use another unused disk )
# Get device path of ROOT DISK
system host-show controller-1 --nowrap | fgrep rootfs
# Get UUID of ROOT DISK by listing disks
system host-disk-list controller-1
# Create new PARTITION on ROOT DISK, and take note of new partitions uuid in response
# Use a partition size such that youll be able to increase docker fs size from 30G to 60G
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol controller-1 <root-disk-uuid> ${PARTITION_SIZE}
# Add new partition to cgts-vg local volume group
system host-pv-add controller-1 cgts-vg <NEW_PARTITION_UUID>
sleep 2 # wait for partition to be added
# Increase docker filesystem to 60G
system host-fs-modify controller-1 docker=60
#. **For OpenStack only:** Configure the system setting for the vSwitch.
.. only:: starlingx
StarlingX has |OVS| (kernel-based) vSwitch configured as default:
* Runs in a container; defined within the helm charts of stx-openstack
manifest.
* Shares the core(s) assigned to the platform.
If you require better performance, |OVS|-|DPDK| (|OVS| with the Data Plane
Development Kit, which is supported only on bare metal hardware) should be
used:
If you require better performance, |OVS-DPDK| (|OVS| with the Data
Plane Development Kit, which is supported only on bare metal hardware)
should be used:
* Runs directly on the host (it is not containerized).
* Requires that at least 1 core be assigned/dedicated to the vSwitch function.
Requires that at least 1 core be assigned/dedicated to the vSwitch
function.
**To deploy the default containerized OVS:**
To deploy the default containerized |OVS|:
::
system modify --vswitch_type none
This does not run any vSwitch directly on the host, instead, it uses the
containerized |OVS| defined in the helm charts of stx-openstack
manifest.
This does not run any vSwitch directly on the host, instead, it uses
the containerized |OVS| defined in the helm charts of
|prefix|-openstack manifest.
**To deploy OVS-DPDK, run the following command:**
To deploy |OVS-DPDK|, run the following command:
::
.. parsed-literal::
system modify --vswitch_type ovs-dpdk
system modify --vswitch_type |ovs-dpdk|
Default recommendation for an |AIO|-controller is to use a single core
for |OVS|-|DPDK| vswitch.
::
# assign 1 core on processor/numa-node 0 on controller-0 to vswitch
system host-cpu-modify -f vswitch -p0 1 controller-0
When using |OVS|-|DPDK|, configure 1x 1G huge page for vSwitch memory on
each |NUMA| node where vswitch is running on this host, with the
following command:
for |OVS-DPDK| vswitch.
.. code-block:: bash
# assign 1x 1G huge page on processor/numa-node 0 on controller-0 to vswitch
# assign 1 core on processor/numa-node 0 on controller-0 to vswitch
system host-cpu-modify -f vswitch -p0 0 controller-0
When using |OVS-DPDK|, configure 1G of huge pages for vSwitch memory on
each |NUMA| node where vswitch is running on the host. It is recommended
to configure 1x 1G huge page (-1G 1) for vSwitch memory on each |NUMA|
node where vswitch is running on host. However, due to a limitation with
Kubernetes, only a single huge page size is supported on any one host. If
your application |VMs| require 2M huge pages, then configure 500x 2M huge
pages (-2M 500) for vSwitch memory on each |NUMA| node where vswitch is
running on host.
.. code-block::
# Assign 1x 1G huge page on processor/numa-node 0 on controller-0 to vswitch
system host-memory-modify -f vswitch -1G 1 controller-0 0
.. important::
|VMs| created in an |OVS|-|DPDK| environment must be configured to use
|VMs| created in an |OVS-DPDK| environment must be configured to use
huge pages to enable networking and must use a flavor with property:
hw:mem_page_size=large
Configure the huge pages for |VMs| in an |OVS|-|DPDK| environment on
this host with the commands:
Configure the huge pages for VMs in an |OVS-DPDK| environment on this
host, assuming 1G huge page size is being used on this host, with the
following commands:
.. code-block:: bash
@ -335,6 +399,7 @@ The newly installed controller needs to be configured.
.. code-block:: bash
# Create nova-local local volume group
system host-lvg-add ${NODE} nova-local
# Get UUID of DISK to create PARTITION to be added to nova-local local volume group
@ -347,7 +412,14 @@ The newly installed controller needs to be configured.
# system host-show ${NODE} --nowrap | fgrep rootfs )
# Create new PARTITION on selected disk, and take note of new partitions uuid in response
PARTITION_SIZE=34 # Use default of 34G for this nova-local partition
# The size of the PARTITION needs to be large enough to hold the aggregate size of
# all nova ephemeral disks of all VMs that you want to be able to host on this host,
# but is limited by the size and space available on the physical disk you chose above.
# The following example uses a small PARTITION size such that you can fit it on the
# root disk, if that is what you chose above.
# Additional PARTITION(s) from additional disks can be added later if required.
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol ${NODE} <disk-uuid> ${PARTITION_SIZE}
# Add new partition to nova-local local volume group
@ -362,7 +434,8 @@ The newly installed controller needs to be configured.
.. important::
A compute-labeled worker host **MUST** have at least one Data class interface.
A compute-labeled |AIO|-controller host **MUST** have at least one
Data class interface.
* Configure the data interfaces for controller-0.
@ -431,7 +504,8 @@ Optionally Configure PCI-SRIOV Interfaces
system host-if-modify -m 1500 -n sriov0 -c pci-sriov ${NODE} <sriov0-if-uuid>
system host-if-modify -m 1500 -n sriov1 -c pci-sriov ${NODE} <sriov1-if-uuid>
# Create Data Networks that the 'pci-sriov' interfaces will be connected to
# If not already created, create Data Networks that the 'pci-sriov' interfaces will
# be connected to
DATANET0='datanet0'
DATANET1='datanet1'
system datanetwork-add ${DATANET0} vlan
@ -442,8 +516,8 @@ Optionally Configure PCI-SRIOV Interfaces
system interface-datanetwork-assign ${NODE} <sriov1-if-uuid> ${DATANET1}
* To enable using |SRIOV| network attachments for the above interfaces in
Kubernetes hosted application containers:
* **For Kubernetes Only:** To enable using |SRIOV| network attachments for
the above interfaces in Kubernetes hosted application containers:
* Configure the Kubernetes |SRIOV| device plugin.

View File

@ -226,8 +226,9 @@ Configure controller-0
#. Configure the |OAM| interface of controller-0 and specify the
attached network as "oam".
Use the |OAM| port name that is applicable to your deployment environment,
for example eth0:
The following example configures the |OAM| interface on a physical untagged
ethernet port, use the |OAM| port name that is applicable to your deployment
environment, for example eth0:
.. code-block:: bash
@ -235,11 +236,15 @@ Configure controller-0
system host-if-modify controller-0 $OAM_IF -c platform
system interface-network-assign controller-0 $OAM_IF oam
To configure a vlan or aggregated ethernet interface, see :ref:`Node
Interfaces <node-interfaces-index>`.
#. Configure the MGMT interface of controller-0 and specify the attached
networks of both "mgmt" and "cluster-host".
Use the MGMT port name that is applicable to your deployment environment,
for example eth1:
The following example configures the MGMT interface on a physical untagged
ethernet port, use the MGMT port name that is applicable to your deployment
environment, for example eth1:
.. code-block:: bash
@ -258,6 +263,8 @@ Configure controller-0
system interface-network-assign controller-0 $MGMT_IF mgmt
system interface-network-assign controller-0 $MGMT_IF cluster-host
To configure a vlan or aggregated ethernet interface, see :ref:`Node
Interfaces <node-interfaces-index>`.
#. Configure |NTP| servers for network time synchronization:
@ -265,6 +272,9 @@ Configure controller-0
system ntp-modify ntpservers=0.pool.ntp.org,1.pool.ntp.org
To configure |PTP| instead of |NTP|, see :ref:`PTP Server Configuration
<ptp-server-config-index>`.
#. If required, configure Ceph storage backend:
A persistent storage backend is required if your application requires |PVCs|.
@ -287,8 +297,8 @@ Configure controller-0
.. important::
**These steps are required only if the StarlingX OpenStack application
(stx-openstack) will be installed.**
These steps are required only if the |prod-os| application
(|prefix|-openstack) will be installed.
#. **For OpenStack only:** Assign OpenStack host labels to controller-0 in
support of installing the stx-openstack manifest and helm-charts later.
@ -299,44 +309,47 @@ Configure controller-0
#. **For OpenStack only:** Configure the system setting for the vSwitch.
.. only:: starlingx
StarlingX has |OVS| (kernel-based) vSwitch configured as default:
* Runs in a container; defined within the helm charts of stx-openstack
manifest.
* Shares the core(s) assigned to the platform.
If you require better performance, |OVS|-|DPDK| (OVS with the Data Plane
Development Kit, which is supported only on bare metal hardware) should
be used:
If you require better performance, |OVS-DPDK| (|OVS| with the Data
Plane Development Kit, which is supported only on bare metal hardware)
should be used:
* Runs directly on the host (it is not containerized).
* Requires that at least 1 core be assigned/dedicated to the vSwitch function.
Requires that at least 1 core be assigned/dedicated to the vSwitch
function.
**To deploy the default containerized OVS:**
To deploy the default containerized |OVS|:
::
system modify --vswitch_type none
This does not run any vSwitch directly on the host, instead, it uses the
containerized |OVS| defined in the helm charts of stx-openstack manifest.
This does not run any vSwitch directly on the host, instead, it uses
the containerized |OVS| defined in the helm charts of stx-openstack
manifest.
**To deploy OVS-DPDK, run the following command:**
To deploy |OVS-DPDK|, run the following command:
::
.. parsed-literal::
system modify --vswitch_type ovs-dpdk
system modify --vswitch_type |ovs-dpdk|
Once vswitch_type is set to OVS-|DPDK|, any subsequent |AIO|-controller
Once vswitch_type is set to |OVS-DPDK|, any subsequent |AIO|-controller
or worker nodes created will default to automatically assigning 1 vSwitch
core for |AIO| controllers and 2 vSwitch cores for compute-labeled worker
nodes.
core for |AIO| controllers and 2 vSwitch cores (1 on each numa-node)
for compute-labeled worker nodes.
.. note::
After controller-0 is unlocked, changing vswitch_type requires
locking and unlocking all compute-labeled worker nodes (and/or |AIO|
controllers) to apply the change.
locking and unlocking controller-0 to apply the change.
.. incl-config-controller-0-storage-end:
@ -436,8 +449,9 @@ Configure controller-1
#. Configure the |OAM| interface of controller-1 and specify the
attached network of "oam".
Use the |OAM| port name that is applicable to your deployment environment,
for example eth0:
The following example configures the |OAM| interface on a physical untagged
ethernet port, use the |OAM| port name that is applicable to your deployment
environment, for example eth0:
.. code-block:: bash
@ -445,6 +459,10 @@ Configure controller-1
system host-if-modify controller-1 $OAM_IF -c platform
system interface-network-assign controller-1 $OAM_IF oam
To configure a vlan or aggregated ethernet interface, see :ref:`Node
Interfaces <node-interfaces-index>`.
#. The MGMT interface is partially set up by the network install procedure;
configuring the port used for network install as the MGMT port and
specifying the attached network of "mgmt".
@ -465,8 +483,8 @@ Configure controller-1
.. important::
**This step is required only if the StarlingX OpenStack application
(stx-openstack) will be installed.**
This step is required only if the |prod-os| application
(|prefix|-openstack) will be installed.
**For OpenStack only:** Assign OpenStack host labels to controller-1 in
support of installing the stx-openstack manifest and helm-charts later.
@ -544,8 +562,8 @@ Configure worker nodes
.. important::
**These steps are required only if the StarlingX OpenStack application
(stx-openstack) will be installed.**
These steps are required only if the |prod-os| application
(|prefix|-openstack) will be installed.
#. **For OpenStack only:** Assign OpenStack host labels to the worker nodes in
support of installing the stx-openstack manifest and helm-charts later.
@ -560,10 +578,10 @@ Configure worker nodes
#. **For OpenStack only:** Configure the host settings for the vSwitch.
**If using OVS-DPDK vswitch, run the following commands:**
If using |OVS-DPDK| vswitch, run the following commands:
Default recommendation for worker node is to use a single core on each
numa-node for |OVS|-|DPDK| vswitch. This should have been automatically
numa-node for |OVS-DPDK| vswitch. This should have been automatically
configured, if not run the following command.
.. code-block:: bash
@ -579,9 +597,15 @@ Configure worker nodes
done
When using |OVS|-|DPDK|, configure 1x 1G huge page for vSwitch memory on
each |NUMA| node where vswitch is running on this host, with the
following command:
When using |OVS-DPDK|, configure 1G of huge pages for vSwitch memory on
each |NUMA| node where vswitch is running on the host. It is recommended
to configure 1x 1G huge page (-1G 1) for vSwitch memory on each |NUMA|
node where vswitch is running on host.
However, due to a limitation with Kubernetes, only a single huge page
size is supported on any one host. If your application |VMs| require 2M
huge pages, then configure 500x 2M huge pages (-2M 500) for vSwitch
memory on each |NUMA| node where vswitch is running on host.
.. code-block:: bash
@ -598,12 +622,13 @@ Configure worker nodes
.. important::
|VMs| created in an |OVS|-|DPDK| environment must be configured to use
huge pages to enable networking and must use a flavor with property:
hw:mem_page_size=large
|VMs| created in an |OVS-DPDK| environment must be configured to use
huge pages to enable networking and must use a flavor with the
property ``hw:mem_page_size=large``
Configure the huge pages for |VMs| in an |OVS|-|DPDK| environment for
this host with the command:
Configure the huge pages for |VMs| in an |OVS-DPDK| environment on
this host, assuming 1G huge page size is being used on this host, with
the following commands:
.. code-block:: bash
@ -635,7 +660,14 @@ Configure worker nodes
# system host-show ${NODE} --nowrap | fgrep rootfs )
# Create new PARTITION on selected disk, and take note of new partitions uuid in response
PARTITION_SIZE=34 # Use default of 34G for this nova-local partition
# The size of the PARTITION needs to be large enough to hold the aggregate size of
# all nova ephemeral disks of all VMs that you want to be able to host on this host,
# but is limited by the size and space available on the physical disk you chose above.
# The following example uses a small PARTITION size such that you can fit it on the
# root disk, if that is what you chose above.
# Additional PARTITION(s) from additional disks can be added later if required.
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol ${NODE} <disk-uuid> ${PARTITION_SIZE}
# Add new partition to nova-local local volume group
@ -645,12 +677,13 @@ Configure worker nodes
#. **For OpenStack only:** Configure data interfaces for worker nodes.
Data class interfaces are vswitch interfaces used by vswitch to provide
VM virtio vNIC connectivity to OpenStack Neutron Tenant Networks on the
|VM| virtio vNIC connectivity to OpenStack Neutron Tenant Networks on the
underlying assigned Data Network.
.. important::
A compute-labeled worker host **MUST** have at least one Data class interface.
A compute-labeled worker host **MUST** have at least one Data class
interface.
* Configure the data interfaces for worker nodes.
@ -697,7 +730,7 @@ Optionally Configure PCI-SRIOV Interfaces
.. only:: openstack
This step is **optional** for OpenStack. Do this step if using |SRIOV|
vNICs in hosted application VMs. Note that pci-sriov interfaces can
vNICs in hosted application |VMs|. Note that pci-sriov interfaces can
have the same Data Networks assigned to them as vswitch data interfaces.
@ -724,7 +757,8 @@ Optionally Configure PCI-SRIOV Interfaces
system host-if-modify -m 1500 -n sriov0 -c pci-sriov ${NODE} <sriov0-if-uuid>
system host-if-modify -m 1500 -n sriov1 -c pci-sriov ${NODE} <sriov1-if-uuid>
# Create Data Networks that the 'pci-sriov' interfaces will be connected to
# If not already created, create Data Networks that the 'pci-sriov'
# interfaces will be connected to
DATANET0='datanet0'
DATANET1='datanet1'
system datanetwork-add ${DATANET0} vlan
@ -735,8 +769,8 @@ Optionally Configure PCI-SRIOV Interfaces
system interface-datanetwork-assign ${NODE} <sriov1-if-uuid> ${DATANET1}
* To enable using |SRIOV| network attachments for the above interfaces in
Kubernetes hosted application containers:
* **For Kubernetes only:** To enable using |SRIOV| network attachments for
the above interfaces in Kubernetes hosted application containers:
* Configure the Kubernetes |SRIOV| device plugin.

View File

@ -273,8 +273,8 @@ Configure worker nodes
.. important::
**These steps are required only if the StarlingX OpenStack application
(stx-openstack) will be installed.**
These steps are required only if the |prod-os| application
(|prefix|-openstack) will be installed.
#. **For OpenStack only:** Assign OpenStack host labels to the worker nodes in
support of installing the stx-openstack manifest and helm-charts later.
@ -289,10 +289,10 @@ Configure worker nodes
#. **For OpenStack only:** Configure the host settings for the vSwitch.
**If using OVS-DPDK vswitch, run the following commands:**
If using |OVS-DPDK| vswitch, run the following commands:
Default recommendation for worker node is to use a single core on each
numa-node for |OVS|-|DPDK| vswitch. This should have been automatically
numa-node for |OVS-DPDK| vswitch. This should have been automatically
configured, if not run the following command.
.. code-block:: bash
@ -308,9 +308,15 @@ Configure worker nodes
done
When using |OVS|-|DPDK|, configure 1x 1G huge page for vSwitch memory on
each |NUMA| node where vswitch is running on this host, with the
following command:
When using |OVS-DPDK|, configure 1G of huge pages for vSwitch memory on
each |NUMA| node where vswitch is running on the host. It is recommended
to configure 1x 1G huge page (-1G 1) for vSwitch memory on each |NUMA|
node where vswitch is running on host.
However, due to a limitation with Kubernetes, only a single huge page
size is supported on any one host. If your application |VMs| require 2M
huge pages, then configure 500x 2M huge pages (-2M 500) for vSwitch
memory on each |NUMA| node where vswitch is running on host.
.. code-block:: bash
@ -327,12 +333,13 @@ Configure worker nodes
.. important::
|VMs| created in an |OVS|-|DPDK| environment must be configured to use
|VMs| created in an |OVS-DPDK| environment must be configured to use
huge pages to enable networking and must use a flavor with property:
hw:mem_page_size=large
Configure the huge pages for |VMs| in an |OVS|-|DPDK| environment for
this host with the command:
Configure the huge pages for |VMs| in an |OVS-DPDK| environment on
this host, assuming 1G huge page size is being used on this host, with
the following commands:
.. code-block:: bash
@ -364,7 +371,14 @@ Configure worker nodes
# system host-show ${NODE} --nowrap | fgrep rootfs )
# Create new PARTITION on selected disk, and take note of new partitions uuid in response
PARTITION_SIZE=34 # Use default of 34G for this nova-local partition
# The size of the PARTITION needs to be large enough to hold the aggregate size of
# all nova ephemeral disks of all VMs that you want to be able to host on this host,
# but is limited by the size and space available on the physical disk you chose above.
# The following example uses a small PARTITION size such that you can fit it on the
# root disk, if that is what you chose above.
# Additional PARTITION(s) from additional disks can be added later if required.
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol ${NODE} <disk-uuid> ${PARTITION_SIZE}
# Add new partition to nova-local local volume group
@ -374,12 +388,13 @@ Configure worker nodes
#. **For OpenStack only:** Configure data interfaces for worker nodes.
Data class interfaces are vswitch interfaces used by vswitch to provide
VM virtio vNIC connectivity to OpenStack Neutron Tenant Networks on the
|VM| virtio vNIC connectivity to OpenStack Neutron Tenant Networks on the
underlying assigned Data Network.
.. important::
A compute-labeled worker host **MUST** have at least one Data class interface.
A compute-labeled worker host **MUST** have at least one Data class
interface.
* Configure the data interfaces for worker nodes.
@ -426,7 +441,7 @@ Optionally Configure PCI-SRIOV Interfaces
.. only:: openstack
This step is **optional** for OpenStack. Do this step if using |SRIOV|
vNICs in hosted application VMs. Note that pci-sriov interfaces can
vNICs in hosted application |VMs|. Note that pci-sriov interfaces can
have the same Data Networks assigned to them as vswitch data interfaces.
@ -453,7 +468,8 @@ Optionally Configure PCI-SRIOV Interfaces
system host-if-modify -m 1500 -n sriov0 -c pci-sriov ${NODE} <sriov0-if-uuid>
system host-if-modify -m 1500 -n sriov1 -c pci-sriov ${NODE} <sriov1-if-uuid>
# Create Data Networks that the 'pci-sriov' interfaces will be connected to
# If not created already, create Data Networks that the 'pci-sriov'
# interfaces will be connected to
DATANET0='datanet0'
DATANET1='datanet1'
system datanetwork-add ${DATANET0} vlan
@ -464,8 +480,8 @@ Optionally Configure PCI-SRIOV Interfaces
system interface-datanetwork-assign ${NODE} <sriov1-if-uuid> ${DATANET1}
* To enable using |SRIOV| network attachments for the above interfaces in
Kubernetes hosted application containers:
* **For Kubernetes only:** To enable using |SRIOV| network attachments for
the above interfaces in Kubernetes hosted application containers:
* Configure the Kubernetes |SRIOV| device plugin.

View File

@ -17,18 +17,21 @@
#. **For OpenStack only:** Configure the system setting for the vSwitch.
.. only:: starlingx
StarlingX has |OVS| (kernel-based) vSwitch configured as default:
* Runs in a container; defined within the helm charts of stx-openstack
manifest.
* Shares the core(s) assigned to the platform.
If you require better performance, |OVS|-|DPDK| (|OVS| with the Data Plane
Development Kit, which is supported only on bare metal hardware) should be
used:
If you require better performance, |OVS-DPDK| (|OVS| with the Data
Plane Development Kit, which is supported only on bare metal hardware)
should be used:
* Runs directly on the host (it is not containerized).
* Requires that at least 1 core be assigned/dedicated to the vSwitch function.
Requires that at least 1 core be assigned/dedicated to the vSwitch
function.
To deploy the default containerized |OVS|:
@ -36,39 +39,40 @@
system modify --vswitch_type none
Do not run any vSwitch directly on the host, instead, use the containerized
|OVS| defined in the helm charts of stx-openstack manifest.
This does not run any vSwitch directly on the host, instead, it uses
the containerized |OVS| defined in the helm charts of stx-openstack
manifest.
To deploy |OVS|-|DPDK|, run the following command:
To deploy |OVS-DPDK|, run the following command:
::
.. parsed-literal::
system modify --vswitch_type ovs-dpdk
system host-cpu-modify -f vswitch -p0 1 controller-0
system modify --vswitch_type |ovs-dpdk|
Once vswitch_type is set to |OVS|-|DPDK|, any subsequent nodes created
will default to automatically assigning 1 vSwitch core for |AIO|
controllers and 2 vSwitch cores for compute-labeled worker nodes.
Default recommendation for an |AIO|-controller is to use a single core
for |OVS-DPDK| vswitch.
When using |OVS|-|DPDK|, configure vSwitch memory per |NUMA| node with
the following command:
.. code-block:: bash
::
# assign 1 core on processor/numa-node 0 on controller-0 to vswitch
system host-cpu-modify -f vswitch -p0 0 controller-0
system host-memory-modify -f <function> -1G <1G hugepages number> <hostname or id> <processor>
For example:
When using |OVS-DPDK|, configure 1x 1G huge page for vSwitch memory on
each |NUMA| node where vswitch is running on this host, with the
following command:
::
.. code-block:: bash
system host-memory-modify -f vswitch -1G 1 worker-0 0
# assign 1x 1G huge page on processor/numa-node 0 on controller-0 to vswitch
system host-memory-modify -f vswitch -1G 1 controller-0 0
|VMs| created in an |OVS|-|DPDK| environment must be configured to use
|VMs| created in an |OVS-DPDK| environment must be configured to use
huge pages to enable networking and must use a flavor with property:
hw:mem_page_size=large
Configure the huge pages for |VMs| in an |OVS|-|DPDK| environment with
the command:
Configure the huge pages for |VMs| in an |OVS-DPDK| environment with the
command:
::
@ -78,7 +82,8 @@
::
system host-memory-modify worker-0 0 -1G 10
system host-memory-modify controller-0 0 -1G 10
system host-memory-modify controller-0 1 -1G 10
.. note::

View File

@ -261,12 +261,11 @@ OpenStack-specific host configuration
system host-cpu-modify -f vswitch -p0 1 controller-0
Once vswitch_type is set to |OVS|-|DPDK|, any subsequent nodes created will
default to automatically assigning 1 vSwitch core for AIO controllers and 2
vSwitch cores for compute-labeled worker nodes.
default to automatically assigning 1 vSwitch core for |AIO| controllers and
2 vSwitch cores for compute-labeled worker nodes.
When using |OVS|-|DPDK|, configure vSwitch memory per NUMA node with the
following
command:
When using |OVS|-|DPDK|, configure vSwitch memory per |NUMA| node with the
following command:
::
@ -333,7 +332,8 @@ Unlock controller-0 in order to bring it into service:
system host-unlock controller-0
Controller-0 will reboot in order to apply configuration changes and come into
service. This can take 5-10 minutes, depending on the performance of the host machine.
service. This can take 5-10 minutes, depending on the performance of the host
machine.
-------------------------------------------------
Install software on controller-1 and worker nodes
@ -377,17 +377,18 @@ Install software on controller-1 and worker nodes
system host-update 3 personality=worker hostname=worker-0
Repeat for worker-1. Power on worker-1 and wait for the new host (hostname=None) to
be discovered by checking 'system host-list':
Repeat for worker-1. Power on worker-1 and wait for the new host
(hostname=None) to be discovered by checking 'system host-list':
::
system host-update 4 personality=worker hostname=worker-1
For rook storage, there is no storage personality. Some hosts with worker personality
providers storage service. Here we still named these worker host storage-x.
Repeat for storage-0 and storage-1. Power on storage-0, storage-1 and wait for the
new host (hostname=None) to be discovered by checking 'system host-list':
For rook storage, there is no storage personality. Some hosts with worker
personality providers storage service. Here we still named these worker host
storage-x. Repeat for storage-0 and storage-1. Power on storage-0, storage-1
and wait for the new host (hostname=None) to be discovered by checking
'system host-list':
::
@ -401,9 +402,9 @@ Install software on controller-1 and worker nodes
A node with Edgeworker personality is also available. See
:ref:`deploy-edgeworker-nodes` for details.
#. Wait for the software installation on controller-1, worker-0, and worker-1 to
complete, for all servers to reboot, and for all to show as locked/disabled/online in
'system host-list'.
#. Wait for the software installation on controller-1, worker-0, and worker-1
to complete, for all servers to reboot, and for all to show as
locked/disabled/online in 'system host-list'.
::
@ -426,9 +427,9 @@ Configure controller-1
.. incl-config-controller-1-start:
Configure the OAM and MGMT interfaces of controller-0 and specify the attached
networks. Use the OAM and MGMT port names, for example eth0, that are applicable
to your deployment environment.
Configure the |OAM| and MGMT interfaces of controller-0 and specify the
attached networks. Use the |OAM| and MGMT port names, for example eth0, that
are applicable to your deployment environment.
(Note that the MGMT interface is partially set up automatically by the network
install procedure.)
@ -516,12 +517,12 @@ Configure worker nodes
This step is **required** for OpenStack.
This step is optional for Kubernetes: Do this step if using SRIOV network
attachments in hosted application containers.
This step is optional for Kubernetes: Do this step if using |SRIOV|
network attachments in hosted application containers.
For Kubernetes SRIOV network attachments:
For Kubernetes |SRIOV| network attachments:
* Configure SRIOV device plug in:
* Configure |SRIOV| device plug in:
::
@ -529,10 +530,10 @@ Configure worker nodes
system host-label-assign ${NODE} sriovdp=enabled
done
* If planning on running DPDK in containers on this host, configure the number
of 1G Huge pages required on both NUMA nodes:
* If planning on running |DPDK| in containers on this host, configure the
number of 1G Huge pages required on both |NUMA| nodes:
::
.. code-block:: bash
for NODE in worker-0 worker-1; do
system host-memory-modify ${NODE} 0 -1G 100
@ -541,7 +542,7 @@ Configure worker nodes
For both Kubernetes and OpenStack:
::
.. code-block:: bash
DATA0IF=<DATA-0-PORT>
DATA1IF=<DATA-1-PORT>
@ -587,7 +588,7 @@ OpenStack-specific host configuration
#. **For OpenStack only:** Assign OpenStack host labels to the worker nodes in
support of installing the stx-openstack manifest and helm-charts later.
::
.. code-block:: bash
for NODE in worker-0 worker-1; do
system host-label-assign $NODE openstack-compute-node=enabled
@ -598,7 +599,7 @@ OpenStack-specific host configuration
#. **For OpenStack only:** Set up disk partition for nova-local volume group,
which is needed for stx-openstack nova ephemeral disks.
::
.. code-block:: bash
for NODE in worker-0 worker-1; do
echo "Configuring Nova local for: $NODE"
@ -617,14 +618,15 @@ Unlock worker nodes
Unlock worker nodes in order to bring them into service:
::
.. code-block:: bash
for NODE in worker-0 worker-1; do
system host-unlock $NODE
done
The worker nodes will reboot in order to apply configuration changes and come into
service. This can take 5-10 minutes, depending on the performance of the host machine.
The worker nodes will reboot in order to apply configuration changes and come
into service. This can take 5-10 minutes, depending on the performance of the
host machine.
-----------------------
Configure storage nodes
@ -632,9 +634,10 @@ Configure storage nodes
#. Assign the cluster-host network to the MGMT interface for the storage nodes.
Note that the MGMT interfaces are partially set up by the network install procedure.
Note that the MGMT interfaces are partially set up by the network install
procedure.
::
.. code-block:: bash
for NODE in storage-0 storage-1; do
system interface-network-assign $NODE mgmt0 cluster-host
@ -653,7 +656,7 @@ Unlock storage nodes
Unlock storage nodes in order to bring them into service:
::
.. code-block:: bash
for STORAGE in storage-0 storage-1; do
system host-unlock $STORAGE
@ -715,7 +718,7 @@ On host storage-0 and storage-1:
system application-apply rook-ceph-apps
#. Wait for OSDs pod ready.
#. Wait for |OSDs| pod ready.
::

View File

@ -218,6 +218,10 @@ For host-based Ceph,
system host-disk-list controller-0 | awk '/\/dev\/sdb/{print $2}' | xargs -i system host-stor-add controller-0 {}
system host-stor-list controller-0
See :ref:`configure-ceph-osds-on-a-host <configure-ceph-osds-on-a-host>` for
additional info on configuring the Ceph storage backend such as supporting
SSD-backed journals, multiple storage tiers, and so on.
For Rook container-based Ceph:
#. Initialize with add ceph-rook backend:

View File

@ -209,6 +209,10 @@ For host-based Ceph,
system host-disk-list controller-0 | awk '/\/dev\/sdb/{print $2}' | xargs -i system host-stor-add controller-0 {}
system host-stor-list controller-0
See :ref:`configure-ceph-osds-on-a-host <configure-ceph-osds-on-a-host>` for
additional info on configuring the Ceph storage backend such as supporting
SSD-backed journals, multiple storage tiers, and so on.
For Rook container-based Ceph:
#. Initialize with add ceph-rook backend:

View File

@ -36,7 +36,7 @@ Install software on worker nodes
#. Using the host id, set the personality of this host to 'worker':
::
.. code-block:: bash
system host-update 3 personality=worker hostname=worker-0
system host-update 4 personality=worker hostname=worker-1
@ -127,9 +127,15 @@ Configure worker nodes
done
When using |OVS|-|DPDK|, configure 1x 1G huge page for vSwitch memory on
each |NUMA| node where vswitch is running on this host, with the
following command:
When using |OVS|-|DPDK|, configure 1G of huge pages for vSwitch memory on
each |NUMA| node where vswitch is running on the host. It is recommended
to configure 1x 1G huge page (-1G 1) for vSwitch memory on each |NUMA|
node where vswitch is running on host.
However, due to a limitation with Kubernetes, only a single huge page
size is supported on any one host. If your application VMs require 2M
huge pages, then configure 500x 2M huge pages (-2M 500) for vSwitch
memory on each |NUMA| node where vswitch is running on host.
.. code-block:: bash
@ -150,8 +156,9 @@ Configure worker nodes
huge pages to enable networking and must use a flavor with property:
hw:mem_page_size=large
Configure the huge pages for |VMs| in an |OVS|-|DPDK| environment for
this host with the command:
Configure the huge pages for |VMs| in an |OVS-DPDK| environment on
this host, assuming 1G huge page size is being used on this host, with
the following commands:
.. code-block:: bash
@ -183,7 +190,14 @@ Configure worker nodes
# system host-show ${NODE} --nowrap | fgrep rootfs )
# Create new PARTITION on selected disk, and take note of new partitions uuid in response
PARTITION_SIZE=34 # Use default of 34G for this nova-local partition
# The size of the PARTITION needs to be large enough to hold the aggregate size of
# all nova ephemeral disks of all VMs that you want to be able to host on this host,
# but is limited by the size and space available on the physical disk you chose above.
# The following example uses a small PARTITION size such that you can fit it on the
# root disk, if that is what you chose above.
# Additional PARTITION(s) from additional disks can be added later if required.
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol ${NODE} <disk-uuid> ${PARTITION_SIZE}
# Add new partition to nova-local local volume group
@ -272,7 +286,8 @@ Optionally Configure PCI-SRIOV Interfaces
system host-if-modify -m 1500 -n sriov0 -c pci-sriov ${NODE} <sriov0-if-uuid>
system host-if-modify -m 1500 -n sriov1 -c pci-sriov ${NODE} <sriov1-if-uuid>
# Create Data Networks that the 'pci-sriov' interfaces will be connected to
# If not already created, create Data Networks that the 'pci-sriov'
# interfaces will be connected to
DATANET0='datanet0'
DATANET1='datanet1'
system datanetwork-add ${DATANET0} vlan
@ -283,8 +298,8 @@ Optionally Configure PCI-SRIOV Interfaces
system interface-datanetwork-assign ${NODE} <sriov1-if-uuid> ${DATANET1}
* To enable using |SRIOV| network attachments for the above interfaces in
Kubernetes hosted application containers:
* **For Kubernetes only** To enable using |SRIOV| network attachments for
the above interfaces in Kubernetes hosted application containers:
* Configure the Kubernetes |SRIOV| device plugin.
@ -326,4 +341,3 @@ Unlock worker nodes in order to bring them into service:
The worker nodes will reboot to apply configuration changes and come into
service. This can take 5-10 minutes, depending on the performance of the host
machine.

View File

@ -39,7 +39,8 @@ Bootstrap system on controller-0
--------------------------------
#. Login using the username / password of "sysadmin" / "sysadmin".
When logging in for the first time, you will be forced to change the password.
When logging in for the first time, you will be forced to change the
password.
::
@ -107,9 +108,9 @@ Bootstrap system on controller-0
#. Create a minimal user configuration override file.
To use this method, create your override file at ``$HOME/localhost.yml``
and provide the minimum required parameters for the deployment configuration
as shown in the example below. Use the OAM IP SUBNET and IP ADDRESSing
applicable to your deployment environment.
and provide the minimum required parameters for the deployment
configuration as shown in the example below. Use the OAM IP SUBNET and IP
ADDRESSing applicable to your deployment environment.
::
@ -222,8 +223,9 @@ Configure controller-0
#. Configure the |OAM| interface of controller-0 and specify the
attached network as "oam".
Use the |OAM| port name that is applicable to your deployment environment,
for example eth0:
The following example configures the |OAM| interface on a physical untagged
ethernet port. Use the |OAM| port name that is applicable to your
deployment environment, for example eth0:
.. code-block:: bash
@ -231,11 +233,15 @@ Configure controller-0
system host-if-modify controller-0 $OAM_IF -c platform
system interface-network-assign controller-0 $OAM_IF oam
To configure a vlan or aggregated ethernet interface, see :ref:`Node
Interfaces <node-interfaces-index>`.
#. Configure the MGMT interface of controller-0 and specify the attached
networks of both "mgmt" and "cluster-host".
Use the MGMT port name that is applicable to your deployment environment,
for example eth1:
The following example configures the MGMT interface on a physical untagged
ethernet port. Use the MGMT port name that is applicable to your deployment
environment, for example eth1:
.. code-block:: bash
@ -249,12 +255,18 @@ Configure controller-0
system interface-network-assign controller-0 $MGMT_IF mgmt
system interface-network-assign controller-0 $MGMT_IF cluster-host
To configure a vlan or aggregated ethernet interface, see :ref:`Node
Interfaces <node-interfaces-index>`.
#. Configure |NTP| servers for network time synchronization:
::
system ntp-modify ntpservers=0.pool.ntp.org,1.pool.ntp.org
To configure |PTP| instead of |NTP|, see ref:`PTP Server Configuration
<ptp-server-config-index>`.
.. only:: openstack
*************************************
@ -276,56 +288,111 @@ Configure controller-0
system host-label-assign controller-0 openvswitch=enabled
system host-label-assign controller-0 sriov=enabled
#. **For OpenStack only:** Due to the additional openstack services running
on the |AIO| controller platform cores, a minimum of 4 platform cores are
required, 6 platform cores are recommended.
Increase the number of platform cores with the following commands:
.. code-block::
# assign 6 cores on processor/numa-node 0 on controller-0 to platform
system host-cpu-modify -f platform -p0 6 controller-0
#. Due to the additional openstack services containers running on the
controller host, the size of the docker filesystem needs to be
increased from the default size of 30G to 60G.
.. code-block:: bash
# check existing size of docker fs
system host-fs-list controller-1
# check available space (Avail Size (GiB)) in cgts-vg LVG where docker fs is located
system host-lvg-list controller-1
# if existing docker fs size + cgts-vg available space is less than 60G,
# you will need to add a new disk partition to cgts-vg
# Assuming you have unused space on ROOT DISK, add partition to ROOT DISK.
# ( if not use another unused disk )
# Get device path of ROOT DISK
system host-show controller-1 --nowrap | fgrep rootfs
# Get UUID of ROOT DISK by listing disks
system host-disk-list controller-1
# Create new PARTITION on ROOT DISK, and take note of new partitions uuid in response
# Use a partition size such that youll be able to increase docker fs size from 30G to 60G
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol controller-1 <root-disk-uuid> ${PARTITION_SIZE}
# Add new partition to cgts-vg local volume group
system host-pv-add controller-1 cgts-vg <NEW_PARTITION_UUID>
sleep 2 # wait for partition to be added
# Increase docker filesystem to 60G
system host-fs-modify controller-1 docker=60
#. **For OpenStack only:** Configure the system setting for the vSwitch.
.. only:: starlingx
StarlingX has |OVS| (kernel-based) vSwitch configured as default:
* Runs in a container; defined within the helm charts of stx-openstack
manifest.
* Shares the core(s) assigned to the platform.
If you require better performance, |OVS|-|DPDK| (|OVS| with the Data
If you require better performance, |OVS-DPDK| (|OVS| with the Data
Plane Development Kit, which is supported only on bare metal hardware)
should be used:
* Runs directly on the host (it is not containerized).
* Requires that at least 1 core be assigned/dedicated to the vSwitch
Requires that at least 1 core be assigned/dedicated to the vSwitch
function.
**To deploy the default containerized OVS:**
To deploy the default containerized |OVS|:
::
system modify --vswitch_type none
This does not run any vSwitch directly on the host, instead, it uses the
containerized |OVS| defined in the helm charts of stx-openstack
This does not run any vSwitch directly on the host, instead, it uses
the containerized |OVS| defined in the helm charts of stx-openstack
manifest.
**To deploy OVS-DPDK, run the following command:**
To deploy |OVS-DPDK|, run the following command:
::
.. parsed-literal::
system modify --vswitch_type ovs-dpdk
system modify --vswitch_type |ovs-dpdk|
Default recommendation for an |AIO|-controller is to use a single core
for |OVS|-|DPDK| vswitch.
Default recommendation for an |AIO|-controller is to use a single
core for |OVS-DPDK| vswitch.
.. code-block:: bash
# assign 1 core on processor/numa-node 0 on controller-0 to vswitch
system host-cpu-modify -f vswitch -p0 1 controller-0
system host-cpu-modify -f vswitch -p0 0 controller-0
Once vswitch_type is set to |OVS|-|DPDK|, any subsequent nodes created
will default to automatically assigning 1 vSwitch core for |AIO|
controllers and 2 vSwitch cores for compute-labeled worker nodes.
Once vswitch_type is set to |OVS-DPDK|, any subsequent nodes created will
default to automatically assigning 1 vSwitch core for |AIO| controllers
and 2 vSwitch cores (1 on each numa-node) for compute-labeled worker
nodes.
When using |OVS-DPDK|, configure 1G huge page for vSwitch memory on each
|NUMA| node where vswitch is running on this host. It is recommended to
configure 1x 1G huge page (-1G 1) for vSwitch memory on each |NUMA| node
where vswitch is running on host.
However, due to a limitation with Kubernetes, only a single huge page
size is supported on any one host. If your application |VMs| require 2M
huge pages, then configure 500x 2M huge pages (-2M 500) for vSwitch
memory on each |NUMA| node where vswitch is running on host.
When using |OVS|-|DPDK|, configure 1x 1G huge page for vSwitch memory on
each |NUMA| node where vswitch is running on this host, with the
following command:
.. code-block:: bash
.. code-block::
# assign 1x 1G huge page on processor/numa-node 0 on controller-0 to vswitch
system host-memory-modify -f vswitch -1G 1 controller-0 0
@ -333,12 +400,13 @@ Configure controller-0
.. important::
|VMs| created in an |OVS|-|DPDK| environment must be configured to use
|VMs| created in an |OVS-DPDK| environment must be configured to use
huge pages to enable networking and must use a flavor with property:
hw:mem_page_size=large
``hw:mem_page_size=large``
Configure the huge pages for |VMs| in an |OVS|-|DPDK| environment on this host with
the commands:
Configure the huge pages for |VMs| in an |OVS-DPDK| environment on
this host, assuming 1G huge page size is being used on this host, with
the following commands:
.. code-block:: bash
@ -349,6 +417,7 @@ Configure controller-0
# assign 10x 1G huge page on processor/numa-node 1 on controller-0 to applications
system host-memory-modify -f application -1G 10 controller-0 1
.. note::
After controller-0 is unlocked, changing vswitch_type requires
@ -372,7 +441,14 @@ Configure controller-0
# system host-show ${NODE} --nowrap | fgrep rootfs )
# Create new PARTITION on selected disk, and take note of new partitions uuid in response
PARTITION_SIZE=34 # Use default of 34G for this nova-local partition
# The size of the PARTITION needs to be large enough to hold the aggregate size of
# all nova ephemeral disks of all VMs that you want to be able to host on this host,
# but is limited by the size and space available on the physical disk you chose above.
# The following example uses a small PARTITION size such that you can fit it on the
# root disk, if that is what you chose above.
# Additional PARTITION(s) from additional disks can be added later if required.
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol ${NODE} <disk-uuid> ${PARTITION_SIZE}
# Add new partition to nova-local local volume group
@ -381,12 +457,13 @@ Configure controller-0
#. **For OpenStack only:** Configure data interfaces for controller-0.
Data class interfaces are vswitch interfaces used by vswitch to provide
VM virtio vNIC connectivity to OpenStack Neutron Tenant Networks on the
|VM| virtio vNIC connectivity to OpenStack Neutron Tenant Networks on the
underlying assigned Data Network.
.. important::
A compute-labeled All-in-one controller host **MUST** have at least one Data class interface.
A compute-labeled All-in-one controller host **MUST** have at least
one Data class interface.
* Configure the data interfaces for controller-0.
@ -454,7 +531,8 @@ Optionally Configure PCI-SRIOV Interfaces
system host-if-modify -m 1500 -n sriov0 -c pci-sriov ${NODE} <sriov0-if-uuid>
system host-if-modify -m 1500 -n sriov1 -c pci-sriov ${NODE} <sriov1-if-uuid>
# Create Data Networks that the 'pci-sriov' interfaces will be connected to
# If not already created, create Data Networks that the 'pci-sriov'
# interfaces will be connected to
DATANET0='datanet0'
DATANET1='datanet1'
system datanetwork-add ${DATANET0} vlan
@ -465,8 +543,8 @@ Optionally Configure PCI-SRIOV Interfaces
system interface-datanetwork-assign ${NODE} <sriov1-if-uuid> ${DATANET1}
* To enable using |SRIOV| network attachments for the above interfaces in
Kubernetes hosted application containers:
* **For Kubernetes Only:** To enable using |SRIOV| network attachments for
the above interfaces in Kubernetes hosted application containers:
* Configure the Kubernetes |SRIOV| device plugin.
@ -608,8 +686,9 @@ Configure controller-1
#. Configure the |OAM| interface of controller-1 and specify the
attached network of "oam".
Use the |OAM| port name that is applicable to your deployment environment,
for example eth0:
The following example configures the |OAM| interface on a physical untagged
ethernet port, use the |OAM| port name that is applicable to your
deployment environment, for example eth0:
::
@ -617,6 +696,9 @@ Configure controller-1
system host-if-modify controller-1 $OAM_IF -c platform
system interface-network-assign controller-1 $OAM_IF oam
To configure a vlan or aggregated ethernet interface, see :ref:`Node
Interfaces <node-interfaces-index>`.
#. The MGMT interface is partially set up by the network install procedure;
configuring the port used for network install as the MGMT port and
specifying the attached network of "mgmt".
@ -636,11 +718,11 @@ Configure controller-1
.. important::
**These steps are required only if the StarlingX OpenStack application
(stx-openstack) will be installed.**
These steps are required only if the |prod-os| application
(|prefix|-openstack) will be installed.
#. **For OpenStack only:** Assign OpenStack host labels to controller-1 in
support of installing the stx-openstack manifest and helm-charts later.
support of installing the |prefix|-openstack manifest and helm-charts later.
::
@ -649,12 +731,58 @@ Configure controller-1
system host-label-assign controller-1 openvswitch=enabled
system host-label-assign controller-1 sriov=enabled
#. **For OpenStack only:** Due to the additional openstack services running
on the |AIO| controller platform cores, a minimum of 4 platform cores are
required, 6 platform cores are recommended.
Increase the number of platform cores with the following commands:
.. code-block::
# assign 6 cores on processor/numa-node 0 on controller-1 to platform
system host-cpu-modify -f platform -p0 6 controller-1
#. Due to the additional openstack services containers running on the
controller host, the size of the docker filesystem needs to be
increased from the default size of 30G to 60G.
.. code-block:: bash
# check existing size of docker fs
system host-fs-list controller-1
# check available space (Avail Size (GiB)) in cgts-vg LVG where docker fs is located
system host-lvg-list controller-1
# if existing docker fs size + cgts-vg available space is less than 60G,
# you will need to add a new disk partition to cgts-vg
# Assuming you have unused space on ROOT DISK, add partition to ROOT DISK.
# ( if not use another unused disk )
# Get device path of ROOT DISK
system host-show controller-1 --nowrap | fgrep rootfs
# Get UUID of ROOT DISK by listing disks
system host-disk-list controller-1
# Create new PARTITION on ROOT DISK, and take note of new partitions uuid in response
# Use a partition size such that youll be able to increase docker fs size from 30G to 60G
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol controller-1 <root-disk-uuid> ${PARTITION_SIZE}
# Add new partition to cgts-vg local volume group
system host-pv-add controller-1 cgts-vg <NEW_PARTITION_UUID>
sleep 2 # wait for partition to be added
# Increase docker filesystem to 60G
system host-fs-modify controller-1 docker=60
#. **For OpenStack only:** Configure the host settings for the vSwitch.
**If using OVS-DPDK vswitch, run the following commands:**
If using |OVS-DPDK| vswitch, run the following commands:
Default recommendation for an AIO-controller is to use a single core
for |OVS|-|DPDK| vswitch. This should have been automatically configured,
Default recommendation for an |AIO|-controller is to use a single core
for |OVS-DPDK| vswitch. This should have been automatically configured,
if not run the following command.
.. code-block:: bash
@ -663,9 +791,16 @@ Configure controller-1
system host-cpu-modify -f vswitch -p0 1 controller-1
When using |OVS|-|DPDK|, configure 1x 1G huge page for vSwitch memory on
each |NUMA| node where vswitch is running on this host, with the
following command:
When using |OVS-DPDK|, configure 1G of huge pages for vSwitch memory on
each |NUMA| node where vswitch is running on the host. It is recommended
to configure 1x 1G huge page (-1G 1) for vSwitch memory on each |NUMA|
node where vswitch is running on host.
However, due to a limitation with Kubernetes, only a single huge page
size is supported on any one host. If your application VMs require 2M
huge pages, then configure 500x 2M huge pages (-2M 500) for vSwitch
memory on each |NUMA| node where vswitch is running on host.
.. code-block:: bash
@ -675,14 +810,15 @@ Configure controller-1
.. important::
|VMs| created in an |OVS|-|DPDK| environment must be configured to use
|VMs| created in an |OVS-DPDK| environment must be configured to use
huge pages to enable networking and must use a flavor with property:
hw:mem_page_size=large
Configure the huge pages for |VMs| in an |OVS|-|DPDK| environment for
this host with the command:
Configure the huge pages for |VMs| in an |OVS-DPDK| environment on
this host, assuming 1G huge page size is being used on this host, with
the following commands:
::
.. code-block:: bash
# assign 10x 1G huge page on processor/numa-node 0 on controller-1 to applications
system host-memory-modify -f application -1G 10 controller-1 0
@ -709,7 +845,14 @@ Configure controller-1
# system host-show ${NODE} --nowrap | fgrep rootfs )
# Create new PARTITION on selected disk, and take note of new partitions uuid in response
PARTITION_SIZE=34 # Use default of 34G for this nova-local partition
# The size of the PARTITION needs to be large enough to hold the aggregate size of
# all nova ephemeral disks of all VMs that you want to be able to host on this host,
# but is limited by the size and space available on the physical disk you chose above.
# The following example uses a small PARTITION size such that you can fit it on the
# root disk, if that is what you chose above.
# Additional PARTITION(s) from additional disks can be added later if required.
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol ${NODE} <disk-uuid> ${PARTITION_SIZE}
# Add new partition to nova-local local volume group
@ -791,7 +934,8 @@ Optionally Configure PCI-SRIOV Interfaces
system host-if-modify -m 1500 -n sriov0 -c pci-sriov ${NODE} <sriov0-if-uuid>
system host-if-modify -m 1500 -n sriov1 -c pci-sriov ${NODE} <sriov1-if-uuid>
# Create Data Networks that the 'pci-sriov' interfaces will be connected to
# If not already created, create Data Networks that the 'pci-sriov' interfaces
# will be connected to
DATANET0='datanet0'
DATANET1='datanet1'
system datanetwork-add ${DATANET0} vlan
@ -802,12 +946,12 @@ Optionally Configure PCI-SRIOV Interfaces
system interface-datanetwork-assign ${NODE} <sriov1-if-uuid> ${DATANET1}
* To enable using |SRIOV| network attachments for the above interfaces in
Kubernetes hosted application containers:
* **For Kubernetes only:** To enable using |SRIOV| network attachments for
the above interfaces in Kubernetes hosted application containers:
* Configure the Kubernetes |SRIOV| device plugin.
::
.. code-block:: bash
system host-label-assign controller-1 sriovdp=enabled
@ -855,7 +999,7 @@ For host-based Ceph:
#. Assign Rook host labels to controller-1 in support of installing the
rook-ceph-apps manifest/helm-charts later:
::
.. code-block:: bash
system host-label-assign controller-1 ceph-mon-placement=enabled
system host-label-assign controller-1 ceph-mgr-placement=enabled
@ -867,7 +1011,7 @@ Unlock controller-1
Unlock controller-1 in order to bring it into service:
::
.. code-block:: bash
system host-unlock controller-1
@ -902,7 +1046,7 @@ machine.
#. Configure Rook to use /dev/sdb on controller-0 and controller-1 as a ceph
|OSD|.
::
.. code-block:: bash
$ system host-disk-wipe -s --confirm controller-0 /dev/sdb
$ system host-disk-wipe -s --confirm controller-1 /dev/sdb

View File

@ -109,8 +109,8 @@ Bootstrap system on controller-0
To use this method, create your override file at ``$HOME/localhost.yml``
and provide the minimum required parameters for the deployment
configuration as shown in the example below. Use the OAM IP SUBNET and IP
ADDRESSing applicable to your deployment environment.
configuration as shown in the example below. Use the |OAM| IP SUBNET and
IP ADDRESSing applicable to your deployment environment.
::
@ -220,9 +220,10 @@ The newly installed controller needs to be configured.
source /etc/platform/openrc
#. Configure the |OAM| interface of controller-0 and specify the attached network
as "oam". Use the |OAM| port name that is applicable to your deployment
environment, for example eth0:
#. Configure the |OAM| interface of controller-0 and specify the attached
network as "oam". The following example configures the OAM interface on a
physical untagged ethernet port, use |OAM| port name that is applicable to
your deployment environment, for example eth0:
::
@ -230,12 +231,18 @@ The newly installed controller needs to be configured.
system host-if-modify controller-0 $OAM_IF -c platform
system interface-network-assign controller-0 $OAM_IF oam
To configure a vlan or aggregated ethernet interface, see :ref:`Node
Interfaces <node-interfaces-index>`.
#. Configure |NTP| servers for network time synchronization:
::
system ntp-modify ntpservers=0.pool.ntp.org,1.pool.ntp.org
To configure |PTP| instead of |NTP|, see :ref:`PTP Server Configuration
<ptp-server-config-index>`.
.. only:: openstack
*************************************
@ -259,63 +266,121 @@ The newly installed controller needs to be configured.
system host-label-assign controller-0 openvswitch=enabled
system host-label-assign controller-0 sriov=enabled
#. **For OpenStack only:** Due to the additional openstack services running
on the |AIO| controller platform cores, a minimum of 4 platform cores are
required, 6 platform cores are recommended.
Increase the number of platform cores with the following commands:
.. code-block::
# Assign 6 cores on processor/numa-node 0 on controller-0 to platform
system host-cpu-modify -f platform -p0 6 controller-0
#. Due to the additional openstack services containers running on the
controller host, the size of the docker filesystem needs to be
increased from the default size of 30G to 60G.
.. code-block:: bash
# check existing size of docker fs
system host-fs-list controller-1
# check available space (Avail Size (GiB)) in cgts-vg LVG where docker fs is located
system host-lvg-list controller-1
# if existing docker fs size + cgts-vg available space is less than 60G,
# you will need to add a new disk partition to cgts-vg
# Assuming you have unused space on ROOT DISK, add partition to ROOT DISK.
# ( if not use another unused disk )
# Get device path of ROOT DISK
system host-show controller-1 --nowrap | fgrep rootfs
# Get UUID of ROOT DISK by listing disks
system host-disk-list controller-1
# Create new PARTITION on ROOT DISK, and take note of new partitions uuid in response
# Use a partition size such that youll be able to increase docker fs size from 30G to 60G
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol controller-1 <root-disk-uuid> ${PARTITION_SIZE}
# Add new partition to cgts-vg local volume group
system host-pv-add controller-1 cgts-vg <NEW_PARTITION_UUID>
sleep 2 # wait for partition to be added
# Increase docker filesystem to 60G
system host-fs-modify controller-1 docker=60
#. **For OpenStack only:** Configure the system setting for the vSwitch.
.. only:: starlingx
StarlingX has |OVS| (kernel-based) vSwitch configured as default:
* Runs in a container; defined within the helm charts of stx-openstack
manifest.
* Shares the core(s) assigned to the platform.
If you require better performance, |OVS|-|DPDK| (|OVS| with the Data Plane
Development Kit, which is supported only on bare metal hardware) should be
used:
If you require better performance, |OVS-DPDK| (|OVS| with the Data
Plane Development Kit, which is supported only on bare metal hardware)
should be used:
* Runs directly on the host (it is not containerized).
* Requires that at least 1 core be assigned/dedicated to the vSwitch function.
Requires that at least 1 core be assigned/dedicated to the vSwitch
function.
**To deploy the default containerized OVS:**
To deploy the default containerized |OVS|:
::
system modify --vswitch_type none
This does not run any vSwitch directly on the host, instead, it uses the
containerized |OVS| defined in the helm charts of stx-openstack
manifest.
This does not run any vSwitch directly on the host, instead, it uses
the containerized |OVS| defined in the helm charts of
|prefix|-openstack manifest.
**To deploy OVS-DPDK, run the following command:**
To deploy |OVS-DPDK|, run the following command:
::
.. parsed-literal::
system modify --vswitch_type ovs-dpdk
system modify --vswitch_type |ovs-dpdk|
Default recommendation for an AIO-controller is to use a single core
for |OVS|-|DPDK| vswitch.
Default recommendation for an |AIO|-controller is to use a single core
for |OVS-DPDK| vswitch.
::
.. code-block:: bash
# assign 1 core on processor/numa-node 0 on controller-0 to vswitch
system host-cpu-modify -f vswitch -p0 1 controller-0
system host-cpu-modify -f vswitch -p0 0 controller-0
When using |OVS|-|DPDK|, configure 1x 1G huge page for vSwitch memory on each |NUMA| node
where vswitch is running on this host, with the following command:
::
When using |OVS-DPDK|, configure 1G of huge pages for vSwitch memory on
each |NUMA| node where vswitch is running on the host. It is recommended
to configure 1x 1G huge page (-1G 1) for vSwitch memory on each |NUMA|
node where vswitch is running on host. However, due to a limitation with
Kubernetes, only a single huge page size is supported on any one host. If
your application |VMs| require 2M huge pages, then configure 500x 2M huge
pages (-2M 500) for vSwitch memory on each |NUMA| node where vswitch is
running on host.
# assign 1x 1G huge page on processor/numa-node 0 on controller-0 to vswitch
.. code-block::
# Assign 1x 1G huge page on processor/numa-node 0 on controller-0 to vswitch
system host-memory-modify -f vswitch -1G 1 controller-0 0
.. important::
|VMs| created in an |OVS|-|DPDK| environment must be configured to use
|VMs| created in an |OVS-DPDK| environment must be configured to use
huge pages to enable networking and must use a flavor with property:
hw:mem_page_size=large
Configure the huge pages for |VMs| in an |OVS|-|DPDK| environment on this host with
the commands:
Configure the huge pages for VMs in an |OVS-DPDK| environment on this
host, assuming 1G huge page size is being used on this host, with the
following commands:
::
.. code-block:: bash
# assign 1x 1G huge page on processor/numa-node 0 on controller-0 to applications
system host-memory-modify -f application -1G 10 controller-0 0
@ -346,7 +411,14 @@ The newly installed controller needs to be configured.
# system host-show ${NODE} --nowrap | fgrep rootfs )
# Create new PARTITION on selected disk, and take note of new partitions uuid in response
PARTITION_SIZE=34 # Use default of 34G for this nova-local partition
# The size of the PARTITION needs to be large enough to hold the aggregate size of
# all nova ephemeral disks of all VMs that you want to be able to host on this host,
# but is limited by the size and space available on the physical disk you chose above.
# The following example uses a small PARTITION size such that you can fit it on the
# root disk, if that is what you chose above.
# Additional PARTITION(s) from additional disks can be added later if required.
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol ${NODE} <disk-uuid> ${PARTITION_SIZE}
# Add new partition to nova-local local volume group
@ -361,7 +433,8 @@ The newly installed controller needs to be configured.
.. important::
A compute-labeled worker host **MUST** have at least one Data class interface.
A compute-labeled |AIO|-controller host **MUST** have at least one
Data class interface.
* Configure the data interfaces for controller-0.
@ -430,7 +503,8 @@ Optionally Configure PCI-SRIOV Interfaces
system host-if-modify -m 1500 -n sriov0 -c pci-sriov ${NODE} <sriov0-if-uuid>
system host-if-modify -m 1500 -n sriov1 -c pci-sriov ${NODE} <sriov1-if-uuid>
# Create Data Networks that the 'pci-sriov' interfaces will be connected to
# If not already created, create Data Networks that the 'pci-sriov' interfaces will
# be connected to
DATANET0='datanet0'
DATANET1='datanet1'
system datanetwork-add ${DATANET0} vlan
@ -441,8 +515,8 @@ Optionally Configure PCI-SRIOV Interfaces
system interface-datanetwork-assign ${NODE} <sriov1-if-uuid> ${DATANET1}
* To enable using |SRIOV| network attachments for the above interfaces in
Kubernetes hosted application containers:
* **For Kubernetes Only:** To enable using |SRIOV| network attachments for
the above interfaces in Kubernetes hosted application containers:
* Configure the Kubernetes |SRIOV| device plugin.

View File

@ -62,7 +62,7 @@ Bootstrap system on controller-0
PORT, IP-ADDRESS/SUBNET-LENGTH and GATEWAY-IP-ADDRESS applicable to your
deployment environment.
::
.. code-block:: bash
sudo ip address add <IP-ADDRESS>/<SUBNET-LENGTH> dev <PORT>
sudo ip link set up dev <PORT>
@ -111,7 +111,7 @@ Bootstrap system on controller-0
configuration as shown in the example below. Use the OAM IP SUBNET and IP
ADDRESSing applicable to your deployment environment.
::
.. code-block:: bash
cd ~
cat <<EOF > localhost.yml
@ -226,8 +226,9 @@ Configure controller-0
#. Configure the |OAM| interface of controller-0 and specify the
attached network as "oam".
Use the |OAM| port name that is applicable to your deployment environment,
for example eth0:
The following example configures the |OAM| interface on a physical untagged
ethernet port, use the |OAM| port name that is applicable to your deployment
environment, for example eth0:
.. code-block:: bash
@ -235,30 +236,45 @@ Configure controller-0
system host-if-modify controller-0 $OAM_IF -c platform
system interface-network-assign controller-0 $OAM_IF oam
To configure a vlan or aggregated ethernet interface, see :ref:`Node
Interfaces <node-interfaces-index>`.
#. Configure the MGMT interface of controller-0 and specify the attached
networks of both "mgmt" and "cluster-host".
Use the MGMT port name that is applicable to your deployment environment,
for example eth1:
The following example configures the MGMT interface on a physical untagged
ethernet port, use the MGMT port name that is applicable to your deployment
environment, for example eth1:
.. code-block:: bash
MGMT_IF=<MGMT-PORT>
# De-provision loopback interface and
# remove mgmt and cluster-host networks from loopback interface
system host-if-modify controller-0 lo -c none
IFNET_UUIDS=$(system interface-network-list controller-0 | awk '{if ($6=="lo") print $4;}')
for UUID in $IFNET_UUIDS; do
system interface-network-remove ${UUID}
done
# Configure management interface and assign mgmt and cluster-host networks to it
system host-if-modify controller-0 $MGMT_IF -c platform
system interface-network-assign controller-0 $MGMT_IF mgmt
system interface-network-assign controller-0 $MGMT_IF cluster-host
To configure a vlan or aggregated ethernet interface, see :ref:`Node
Interfaces <node-interfaces-index>`.
#. Configure |NTP| servers for network time synchronization:
::
system ntp-modify ntpservers=0.pool.ntp.org,1.pool.ntp.org
To configure |PTP| instead of |NTP|, see :ref:`PTP Server Configuration
<ptp-server-config-index>`.
#. If required, configure Ceph storage backend:
A persistent storage backend is required if your application requires |PVCs|.
@ -281,8 +297,8 @@ Configure controller-0
.. important::
**These steps are required only if the StarlingX OpenStack application
(stx-openstack) will be installed.**
These steps are required only if the |prod-os| application
(|prefix|-openstack) will be installed.
#. **For OpenStack only:** Assign OpenStack host labels to controller-0 in
support of installing the stx-openstack manifest and helm-charts later.
@ -293,44 +309,47 @@ Configure controller-0
#. **For OpenStack only:** Configure the system setting for the vSwitch.
.. only:: starlingx
StarlingX has |OVS| (kernel-based) vSwitch configured as default:
* Runs in a container; defined within the helm charts of stx-openstack
manifest.
* Shares the core(s) assigned to the platform.
If you require better performance, |OVS|-|DPDK| (OVS with the Data Plane
Development Kit, which is supported only on bare metal hardware) should
be used:
If you require better performance, |OVS-DPDK| (|OVS| with the Data
Plane Development Kit, which is supported only on bare metal hardware)
should be used:
* Runs directly on the host (it is not containerized).
* Requires that at least 1 core be assigned/dedicated to the vSwitch function.
Requires that at least 1 core be assigned/dedicated to the vSwitch
function.
**To deploy the default containerized OVS:**
To deploy the default containerized |OVS|:
::
system modify --vswitch_type none
This does not run any vSwitch directly on the host, instead, it uses the
containerized |OVS| defined in the helm charts of stx-openstack manifest.
This does not run any vSwitch directly on the host, instead, it uses
the containerized |OVS| defined in the helm charts of stx-openstack
manifest.
**To deploy OVS-DPDK, run the following command:**
To deploy |OVS-DPDK|, run the following command:
::
.. parsed-literal::
system modify --vswitch_type ovs-dpdk
system modify --vswitch_type |ovs-dpdk|
Once vswitch_type is set to OVS-|DPDK|, any subsequent AIO-controller or
worker nodes created will default to automatically assigning 1 vSwitch
core for |AIO| controllers and 2 vSwitch cores for compute-labeled worker
nodes.
Once vswitch_type is set to |OVS-DPDK|, any subsequent |AIO|-controller
or worker nodes created will default to automatically assigning 1 vSwitch
core for |AIO| controllers and 2 vSwitch cores (1 on each numa-node)
for compute-labeled worker nodes.
.. note::
After controller-0 is unlocked, changing vswitch_type requires
locking and unlocking all compute-labeled worker nodes (and/or |AIO|
controllers) to apply the change.
locking and unlocking controller-0 to apply the change.
.. incl-config-controller-0-storage-end:
@ -430,8 +449,9 @@ Configure controller-1
#. Configure the |OAM| interface of controller-1 and specify the
attached network of "oam".
Use the |OAM| port name that is applicable to your deployment environment,
for example eth0:
The following example configures the |OAM| interface on a physical untagged
ethernet port, use the |OAM| port name that is applicable to your deployment
environment, for example eth0:
.. code-block:: bash
@ -439,6 +459,9 @@ Configure controller-1
system host-if-modify controller-1 $OAM_IF -c platform
system interface-network-assign controller-1 $OAM_IF oam
To configure a vlan or aggregated ethernet interface, see :ref:`Node
Interfaces <node-interfaces-index>`.
#. The MGMT interface is partially set up by the network install procedure;
configuring the port used for network install as the MGMT port and
specifying the attached network of "mgmt".
@ -459,8 +482,8 @@ Configure controller-1
.. important::
**This step is required only if the StarlingX OpenStack application
(stx-openstack) will be installed.**
This step is required only if the |prod-os| application
(|prefix|-openstack) will be installed.
**For OpenStack only:** Assign OpenStack host labels to controller-1 in
support of installing the stx-openstack manifest and helm-charts later.
@ -538,8 +561,8 @@ Configure worker nodes
.. important::
**These steps are required only if the StarlingX OpenStack application
(stx-openstack) will be installed.**
These steps are required only if the |prod-os| application
(|prefix|-openstack) will be installed.
#. **For OpenStack only:** Assign OpenStack host labels to the worker nodes in
support of installing the stx-openstack manifest and helm-charts later.
@ -554,10 +577,10 @@ Configure worker nodes
#. **For OpenStack only:** Configure the host settings for the vSwitch.
**If using OVS-DPDK vswitch, run the following commands:**
If using |OVS-DPDK| vswitch, run the following commands:
Default recommendation for worker node is to use a single core on each
numa-node for |OVS|-|DPDK| vswitch. This should have been automatically
numa-node for |OVS-DPDK| vswitch. This should have been automatically
configured, if not run the following command.
.. code-block:: bash
@ -573,9 +596,15 @@ Configure worker nodes
done
When using |OVS|-|DPDK|, configure 1x 1G huge page for vSwitch memory on
each |NUMA| node where vswitch is running on this host, with the
following command:
When using |OVS-DPDK|, configure 1G of huge pages for vSwitch memory on
each |NUMA| node where vswitch is running on the host. It is recommended
to configure 1x 1G huge page (-1G 1) for vSwitch memory on each |NUMA|
node where vswitch is running on host.
However, due to a limitation with Kubernetes, only a single huge page
size is supported on any one host. If your application |VMs| require 2M
huge pages, then configure 500x 2M huge pages (-2M 500) for vSwitch
memory on each |NUMA| node where vswitch is running on host.
.. code-block:: bash
@ -592,12 +621,13 @@ Configure worker nodes
.. important::
|VMs| created in an |OVS|-|DPDK| environment must be configured to use
huge pages to enable networking and must use a flavor with property:
hw:mem_page_size=large
|VMs| created in an |OVS-DPDK| environment must be configured to use
huge pages to enable networking and must use a flavor with the
property ``hw:mem_page_size=large``
Configure the huge pages for |VMs| in an |OVS|-|DPDK| environment for
this host with the command:
Configure the huge pages for |VMs| in an |OVS-DPDK| environment on
this host, assuming 1G huge page size is being used on this host, with
the following commands:
.. code-block:: bash
@ -629,7 +659,14 @@ Configure worker nodes
# system host-show ${NODE} --nowrap | fgrep rootfs )
# Create new PARTITION on selected disk, and take note of new partitions uuid in response
PARTITION_SIZE=34 # Use default of 34G for this nova-local partition
# The size of the PARTITION needs to be large enough to hold the aggregate size of
# all nova ephemeral disks of all VMs that you want to be able to host on this host,
# but is limited by the size and space available on the physical disk you chose above.
# The following example uses a small PARTITION size such that you can fit it on the
# root disk, if that is what you chose above.
# Additional PARTITION(s) from additional disks can be added later if required.
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol ${NODE} <disk-uuid> ${PARTITION_SIZE}
# Add new partition to nova-local local volume group
@ -639,12 +676,13 @@ Configure worker nodes
#. **For OpenStack only:** Configure data interfaces for worker nodes.
Data class interfaces are vswitch interfaces used by vswitch to provide
VM virtio vNIC connectivity to OpenStack Neutron Tenant Networks on the
|VM| virtio vNIC connectivity to OpenStack Neutron Tenant Networks on the
underlying assigned Data Network.
.. important::
A compute-labeled worker host **MUST** have at least one Data class interface.
A compute-labeled worker host **MUST** have at least one Data class
interface.
* Configure the data interfaces for worker nodes.
@ -718,7 +756,8 @@ Optionally Configure PCI-SRIOV Interfaces
system host-if-modify -m 1500 -n sriov0 -c pci-sriov ${NODE} <sriov0-if-uuid>
system host-if-modify -m 1500 -n sriov1 -c pci-sriov ${NODE} <sriov1-if-uuid>
# Create Data Networks that the 'pci-sriov' interfaces will be connected to
# If not already created, create Data Networks that the 'pci-sriov'
# interfaces will be connected to
DATANET0='datanet0'
DATANET1='datanet1'
system datanetwork-add ${DATANET0} vlan
@ -729,8 +768,8 @@ Optionally Configure PCI-SRIOV Interfaces
system interface-datanetwork-assign ${NODE} <sriov1-if-uuid> ${DATANET1}
* To enable using |SRIOV| network attachments for the above interfaces in
Kubernetes hosted application containers:
* **For Kubernetes only:** To enable using |SRIOV| network attachments for
the above interfaces in Kubernetes hosted application containers:
* Configure the Kubernetes |SRIOV| device plugin.

View File

@ -273,8 +273,8 @@ Configure worker nodes
.. important::
**These steps are required only if the StarlingX OpenStack application
(stx-openstack) will be installed.**
These steps are required only if the |prod-os| application
(|prefix|-openstack) will be installed.
#. **For OpenStack only:** Assign OpenStack host labels to the worker nodes in
support of installing the stx-openstack manifest and helm-charts later.
@ -289,10 +289,10 @@ Configure worker nodes
#. **For OpenStack only:** Configure the host settings for the vSwitch.
**If using OVS-DPDK vswitch, run the following commands:**
If using |OVS-DPDK| vswitch, run the following commands:
Default recommendation for worker node is to use a single core on each
numa-node for |OVS|-|DPDK| vswitch. This should have been automatically
numa-node for |OVS-DPDK| vswitch. This should have been automatically
configured, if not run the following command.
.. code-block:: bash
@ -308,9 +308,15 @@ Configure worker nodes
done
When using |OVS|-|DPDK|, configure 1x 1G huge page for vSwitch memory on
each |NUMA| node where vswitch is running on this host, with the
following command:
When using |OVS-DPDK|, configure 1G of huge pages for vSwitch memory on
each |NUMA| node where vswitch is running on the host. It is recommended
to configure 1x 1G huge page (-1G 1) for vSwitch memory on each |NUMA|
node where vswitch is running on host.
However, due to a limitation with Kubernetes, only a single huge page
size is supported on any one host. If your application |VMs| require 2M
huge pages, then configure 500x 2M huge pages (-2M 500) for vSwitch
memory on each |NUMA| node where vswitch is running on host.
.. code-block:: bash
@ -327,12 +333,13 @@ Configure worker nodes
.. important::
|VMs| created in an |OVS|-|DPDK| environment must be configured to use
|VMs| created in an |OVS-DPDK| environment must be configured to use
huge pages to enable networking and must use a flavor with property:
hw:mem_page_size=large
Configure the huge pages for |VMs| in an |OVS|-|DPDK| environment for
this host with the command:
Configure the huge pages for |VMs| in an |OVS-DPDK| environment on
this host, assuming 1G huge page size is being used on this host, with
the following commands:
.. code-block:: bash
@ -364,7 +371,14 @@ Configure worker nodes
# system host-show ${NODE} --nowrap | fgrep rootfs )
# Create new PARTITION on selected disk, and take note of new partitions uuid in response
PARTITION_SIZE=34 # Use default of 34G for this nova-local partition
# The size of the PARTITION needs to be large enough to hold the aggregate size of
# all nova ephemeral disks of all VMs that you want to be able to host on this host,
# but is limited by the size and space available on the physical disk you chose above.
# The following example uses a small PARTITION size such that you can fit it on the
# root disk, if that is what you chose above.
# Additional PARTITION(s) from additional disks can be added later if required.
PARTITION_SIZE=30
system hostdisk-partition-add -t lvm_phys_vol ${NODE} <disk-uuid> ${PARTITION_SIZE}
# Add new partition to nova-local local volume group
@ -374,12 +388,13 @@ Configure worker nodes
#. **For OpenStack only:** Configure data interfaces for worker nodes.
Data class interfaces are vswitch interfaces used by vswitch to provide
VM virtio vNIC connectivity to OpenStack Neutron Tenant Networks on the
|VM| virtio vNIC connectivity to OpenStack Neutron Tenant Networks on the
underlying assigned Data Network.
.. important::
A compute-labeled worker host **MUST** have at least one Data class interface.
A compute-labeled worker host **MUST** have at least one Data class
interface.
* Configure the data interfaces for worker nodes.
@ -432,7 +447,7 @@ Optionally Configure PCI-SRIOV Interfaces
* Configure the pci-sriov interfaces for worker nodes.
.. code-block:: none
.. code-block:: bash
# Execute the following lines with
export NODE=worker-0
@ -453,7 +468,8 @@ Optionally Configure PCI-SRIOV Interfaces
system host-if-modify -m 1500 -n sriov0 -c pci-sriov ${NODE} <sriov0-if-uuid>
system host-if-modify -m 1500 -n sriov1 -c pci-sriov ${NODE} <sriov1-if-uuid>
# Create Data Networks that the 'pci-sriov' interfaces will be connected to
# If not created already, create Data Networks that the 'pci-sriov'
# interfaces will be connected to
DATANET0='datanet0'
DATANET1='datanet1'
system datanetwork-add ${DATANET0} vlan
@ -464,8 +480,8 @@ Optionally Configure PCI-SRIOV Interfaces
system interface-datanetwork-assign ${NODE} <sriov1-if-uuid> ${DATANET1}
* To enable using |SRIOV| network attachments for the above interfaces in
Kubernetes hosted application containers:
* **For Kubernetes only:** To enable using |SRIOV| network attachments for
the above interfaces in Kubernetes hosted application containers:
* Configure the Kubernetes |SRIOV| device plugin.

View File

@ -17,59 +17,70 @@
#. **For OpenStack only:** Configure the system setting for the vSwitch.
.. only:: starlingx
StarlingX has |OVS| (kernel-based) vSwitch configured as default:
* Runs in a container; defined within the helm charts of stx-openstack
manifest.
* Shares the core(s) assigned to the platform.
If you require better performance, |OVS|-|DPDK| (|OVS| with the Data Plane
Development Kit, which is supported only on bare metal hardware) should be
used:
If you require better performance, |OVS-DPDK| (|OVS| with the Data
Plane Development Kit, which is supported only on bare metal hardware)
should be used:
* Runs directly on the host (it is not containerized).
* Requires that at least 1 core be assigned/dedicated to the vSwitch function.
Requires that at least 1 core be assigned/dedicated to the vSwitch
function.
**To deploy the default containerized OVS:**
To deploy the default containerized |OVS|:
::
system modify --vswitch_type none
This does not run any vSwitch directly on the host, instead, it uses the containerized
|OVS| defined in the helm charts of stx-openstack manifest.
This does not run any vSwitch directly on the host, instead, it uses
the containerized |OVS| defined in the helm charts of stx-openstack
manifest.
**To deploy OVS-DPDK, run the following command:**
To deploy |OVS-DPDK|, run the following command:
::
.. parsed-literal::
system modify --vswitch_type ovs-dpdk
system host-cpu-modify -f vswitch -p0 1 controller-0
system modify --vswitch_type |ovs-dpdk|
Once vswitch_type is set to |OVS|-|DPDK|, any subsequent nodes created
will default to automatically assigning 1 vSwitch core for |AIO|
controllers and 2 vSwitch cores for compute-labeled worker nodes.
Default recommendation for an |AIO|-controller is to use a single core for
|OVS-DPDK| vswitch.
When using |OVS|-|DPDK|, configure vSwitch memory per |NUMA| node with
the following command:
.. code-block:: bash
::
# assign 1 core on processor/numa-node 0 on controller-0 to vswitch
system host-cpu-modify -f vswitch -p0 0 controller-0
When using |OVS-DPDK|, configure 1x 1G huge page for vSwitch memory on
each |NUMA| node where vswitch is running on this host, with the
following command:
.. code-block:: bash
# assign 1x 1G huge page on processor/numa-node 0 on controller-0 to vswitch
system host-memory-modify -f vswitch -1G 1 controller-0 0
system host-memory-modify -f vswitch -1G 1 worker-0 0
.. important::
|VMs| created in an |OVS|-|DPDK| environment must be configured to use
|VMs| created in an |OVS-DPDK| environment must be configured to use
huge pages to enable networking and must use a flavor with property:
hw:mem_page_size=large
Configure the huge pages for |VMs| in an |OVS|-|DPDK| environment with
Configure the huge pages for |VMs| in an |OVS-DPDK| environment with
the command:
::
system host-memory-modify -f application -1G 10 worker-0 0
system host-memory-modify -f application -1G 10 worker-1 1
system host-memory-modify controller-0 0 -1G 10
system host-memory-modify controller-0 1 -1G 10
.. note::

View File

@ -118,7 +118,7 @@ Bootstrap system on controller-0
as shown in the example below. Use the OAM IP SUBNET and IP ADDRESSing
applicable to your deployment environment.
::
.. code-block:: bash
cd ~
cat <<EOF > localhost.yml
@ -180,7 +180,7 @@ Configure controller-0
attached networks. Use the OAM and MGMT port names, for example eth0, that are
applicable to your deployment environment.
::
.. code-block:: bash
OAM_IF=<OAM-PORT>
MGMT_IF=<MGMT-PORT>
@ -242,9 +242,10 @@ OpenStack-specific host configuration
used:
* Runs directly on the host (it is not containerized).
* Requires that at least 1 core be assigned/dedicated to the vSwitch function.
* Requires that at least 1 core be assigned/dedicated to the vSwitch
function.
To deploy the default containerized OVS|:
To deploy the default containerized |OVS|:
::
@ -261,12 +262,11 @@ OpenStack-specific host configuration
system host-cpu-modify -f vswitch -p0 1 controller-0
Once vswitch_type is set to |OVS|-|DPDK|, any subsequent nodes created will
default to automatically assigning 1 vSwitch core for AIO controllers and 2
vSwitch cores for compute-labeled worker nodes.
default to automatically assigning 1 vSwitch core for |AIO| controllers and
2 vSwitch cores for compute-labeled worker nodes.
When using |OVS|-|DPDK|, configure vSwitch memory per NUMA node with the
following
command:
When using |OVS|-|DPDK|, configure vSwitch memory per |NUMA| node with the
following command:
::
@ -282,7 +282,7 @@ OpenStack-specific host configuration
pages to enable networking and must use a flavor with property:
hw:mem_page_size=large
Configure the huge pages for VMs in an |OVS|-|DPDK| environment with the
Configure the huge pages for |VMs| in an |OVS|-|DPDK| environment with the
command:
::
@ -298,7 +298,7 @@ OpenStack-specific host configuration
.. note::
After controller-0 is unlocked, changing vswitch_type requires
locking and unlocking all compute-labeled worker nodes (and/or AIO
locking and unlocking all compute-labeled worker nodes (and/or |AIO|
controllers) to apply the change.
.. incl-config-controller-0-storage-end:
@ -346,8 +346,8 @@ Install software on controller-1 and worker nodes
#. As controller-1 boots, a message appears on its console instructing you to
configure the personality of the node.
#. On the console of controller-0, list hosts to see newly discovered controller-1
host (hostname=None):
#. On the console of controller-0, list hosts to see newly discovered
controller-1 host (hostname=None):
::
@ -428,9 +428,9 @@ Configure controller-1
.. incl-config-controller-1-start:
Configure the |OAM| and MGMT interfaces of controller-0 and specify the attached
networks. Use the |OAM| and MGMT port names, for example eth0, that are applicable
to your deployment environment.
Configure the |OAM| and MGMT interfaces of controller-0 and specify the
attached networks. Use the |OAM| and MGMT port names, for example eth0, that
are applicable to your deployment environment.
(Note that the MGMT interface is partially set up automatically by the network
install procedure.)
@ -518,12 +518,12 @@ Configure worker nodes
This step is **required** for OpenStack.
This step is optional for Kubernetes: Do this step if using SRIOV network
attachments in hosted application containers.
This step is optional for Kubernetes: Do this step if using |SRIOV|
network attachments in hosted application containers.
For Kubernetes SRIOV network attachments:
For Kubernetes |SRIOV| network attachments:
* Configure SRIOV device plug in:
* Configure |SRIOV| device plug in:
::
@ -531,8 +531,8 @@ Configure worker nodes
system host-label-assign ${NODE} sriovdp=enabled
done
* If planning on running |DPDK| in containers on this host, configure the number
of 1G Huge pages required on both |NUMA| nodes:
* If planning on running |DPDK| in containers on this host, configure the
number of 1G Huge pages required on both |NUMA| nodes:
::
@ -619,7 +619,7 @@ Unlock worker nodes
Unlock worker nodes in order to bring them into service:
::
.. code-block:: bash
for NODE in worker-0 worker-1; do
system host-unlock $NODE
@ -638,7 +638,7 @@ Configure storage nodes
Note that the MGMT interfaces are partially set up by the network install
procedure.
::
.. code-block:: bash
for NODE in storage-0 storage-1; do
system interface-network-assign $NODE mgmt0 cluster-host
@ -657,7 +657,7 @@ Unlock storage nodes
Unlock storage nodes in order to bring them into service:
::
.. code-block:: bash
for STORAGE in storage-0 storage-1; do
system host-unlock $STORAGE

View File

@ -218,6 +218,10 @@ For host-based Ceph,
system host-disk-list controller-0 | awk '/\/dev\/sdb/{print $2}' | xargs -i system host-stor-add controller-0 {}
system host-stor-list controller-0
See :ref:`configure-ceph-osds-on-a-host <configure-ceph-osds-on-a-host>` for
additional info on configuring the Ceph storage backend such as supporting
SSD-backed journals, multiple storage tiers, and so on.
For Rook container-based Ceph:
#. Initialize with add ceph-rook backend:

View File

@ -209,6 +209,10 @@ For host-based Ceph,
system host-disk-list controller-0 | awk '/\/dev\/sdb/{print $2}' | xargs -i system host-stor-add controller-0 {}
system host-stor-list controller-0
See :ref:`configure-ceph-osds-on-a-host <configure-ceph-osds-on-a-host>` for
additional info on configuring the Ceph storage backend such as supporting
SSD-backed journals, multiple storage tiers, and so on.
For Rook container-based Ceph:
#. Initialize with add ceph-rook backend:
@ -301,11 +305,11 @@ OpenStack-specific host configuration
#. **For OpenStack only:** A vSwitch is required.
The default vSwitch is containerized OVS that is packaged with the
The default vSwitch is containerized |OVS| that is packaged with the
stx-openstack manifest/helm-charts. StarlingX provides the option to use
OVS-DPDK on the host, however, in the virtual environment OVS-DPDK is NOT
supported, only OVS is supported. Therefore, simply use the default OVS
vSwitch here.
|OVS-DPDK| on the host, however, in the virtual environment |OVS-DPDK| is
NOT supported, only |OVS| is supported. Therefore, simply use the default
|OVS| vSwitch here.
#. **For OpenStack Only:** Set up disk partition for nova-local volume group,
which is needed for stx-openstack nova ephemeral disks.
@ -339,8 +343,8 @@ Unlock virtual controller-0 to bring it into service:
system host-unlock controller-0
Controller-0 will reboot to apply configuration changes and come into
service. This can take 5-10 minutes, depending on the performance of the host machine.
Controller-0 will reboot to apply configuration changes and come into service.
This can take 5-10 minutes, depending on the performance of the host machine.
--------------------------------------------------------------------------
Optionally, finish configuration of Ceph-based Persistent Storage Backend
@ -393,7 +397,7 @@ On **virtual** controller-0:
system application-apply rook-ceph-apps
#. Wait for OSDs pod ready
#. Wait for |OSDs| pod ready.
::

View File

@ -103,6 +103,8 @@ Host memory provisioning
host_memory_provisioning/allocating-host-memory-using-horizon
host_memory_provisioning/allocating-host-memory-using-the-cli
.. _node-interfaces-index:
---------------
Node interfaces
---------------

View File

@ -11,6 +11,12 @@ Install REST API and Horizon Certificate
This certificate must be valid for the domain configured for OpenStack, see the
sections on :ref:`Accessing the System <access-using-the-default-set-up>`.
.. rubric:: |prereq|
Before installing the openstack certificate and key, you must install the ROOT
|CA| for the openstack certificate as a trusted ca, :ref:`Install a Trusted CA
Certificate <install-a-trusted-ca-certificate>`.
.. rubric:: |proc|
#. Install the certificate for OpenStack as Helm chart overrides.
@ -23,7 +29,7 @@ sections on :ref:`Accessing the System <access-using-the-default-set-up>`.
private key.
.. note::
The OpenStack certificate must be created with wildcard SAN.
The OpenStack certificate must be created with wildcard |SAN|.
For example, to create a certificate for |FQDN|: west2.us.example.com,
the following entry must be included in the certificate:

View File

@ -98,6 +98,8 @@ Storage Hosts
storage-hosts-storage-on-storage-hosts
replication-groups
.. _configure-ceph-osds-on-a-host:
-----------------------------
Configure Ceph OSDs on a Host
-----------------------------

View File

@ -35,6 +35,8 @@ NTP Server Configuration
configuring-ntp-servers-and-services-using-the-cli
resynchronizing-a-host-to-the-ntp-server
.. _ptp-server-config-index:
------------------------
PTP Server Configuration
------------------------

View File

@ -18,9 +18,10 @@ hosts are upgraded one at time while continuing to provide its hosting services
to its hosted applications. An upgrade can be performed manually or using
Upgrade Orchestration, which automates much of the upgrade procedure, leaving a
few manual steps to prevent operator oversight. For more information on manual
upgrades, see :ref:`Manual |prod| Components Upgrade
upgrades, see :ref:`Manual PLatform Components Upgrade
<manual-upgrade-overview>`. For more information on upgrade orchestration, see
:ref:`Orchestrated |prod| Component Upgrade <orchestration-upgrade-overview>`.
:ref:`Orchestrated Platform Component Upgrade
<orchestration-upgrade-overview>`.
.. warning::
Do NOT use information in the |updates-doc| guide for |prod-dc|
@ -56,7 +57,7 @@ message formats from the old release of software. Before upgrading the second
controller, is the "point-of-no-return for an in-service abort" of the upgrades
process. The second controller is loaded with the new release of software and
becomes the new Standby controller. For more information on manual upgrades,
see :ref:`Manual |prod| Components Upgrade <manual-upgrade-overview>` .
see :ref:`Manual Platform Components Upgrade <manual-upgrade-overview>` .
If present, storage nodes are locked, upgraded and unlocked one at a time in
order to respect the redundancy model of |prod| storage nodes. Storage nodes