treasuremap/doc/source/airsloop.rst
Kaspars Skels 46616f904f Use HWE kernel for Airsloop computes
Change Airsloop to align with Seaworthy deployment using HWE kernel.

Change-Id: Iad86de5d8381e2bdfb1f6dc23930bda50775beaa
2019-09-14 18:05:44 +00:00

21 KiB

Airsloop: Simple Bare-Metal Airship

Airsloop is a two bare-metal server site deployment reference.

The goal of this site is to be used as a reference for simplified Airship deployments with one control and one or more compute nodes.

It is recommended to get familiar with the Site Authoring and Deployment Guide documentation before deploying Airsloop in the lab. Most steps and concepts including setting up the Genesis node are the same.

image

Various resiliency and security features are tuned down via configuration.

  • Two bare-metal server setup with 1 control, and 1 compute. Most components are scaled to a single replica and doesn't carry any HA as there is only a single control plane host.
  • No requirements for DNS/certificates. HTTP and internal cluster DNS is used.
  • Ceph set to use the single disk. This generally provides minimalistic no-touch Ceph deployment. No replication of Ceph data (single copy).
  • Simplified networking (no bonding). Two network interfaces are used by default (flat PXE, and DATA network with VLANs for OAM, Calico, Storage, and OpenStack Overlay).
  • Generic hostnames used (airsloop-control-1, airsloop-compute-1) that simplifies generation of k8s certificates.

Airsloop site manifests are available at site/airsloop.

Hardware

While HW configuration is flexible, Airsloop reference manifests reflect a single control and a single compute node. The aim of this is to create a minimalistic lab/demo reference environment.

Increasing the number of compute nodes will require site overrides to align parts of the system such as Ceph OSDs, etcd, etc.

See host profiles for the servers here.

Node Hostnames
control airsloop-control-1
compute airsloop-compute-1

Network

Physical (underlay) networks are described in Drydock site configuration here.

It defines OOB (iLO/IPMI), untagged PXE, and multiple tagged general use networks. Also no bonded interfaces are used in Airsloop deployment.

The networking reference is simplified compared to Airship Seaworthy site. There are only two NICs required (excluding oob), one for PXE and another one for the rest of the networks separated using VLAN segmentation.

Below is the reference network configuration:

NICs VLANs Names CIDRs
oob N/A oob 10.22.104.0/24
pxe N/A pxe 10.22.70.0/24

+

data +

71

------------+

72

------------+

73

------------+

74

oam

-----------+

calico

-----------+

storage

-----------+

overlay

10.22.71.0/24 ---------------+ 10.22.72.0/24 ---------------+ 10.22.73.0/24 ---------------+ 10.22.74.0/24

Calico overlay for k8s POD networking uses IPIP mesh.

Storage

Because Airsloop is a minimalistic deployment the required number of disks is just one per node. That disk is not only used by the OS but also by Ceph Journals and OSDs. The way that this is achieved is by using directories and not extra disks for Ceph storage. Ceph OSD configuration can be changed in a Ceph chart override.

The following Ceph chart configuration is used:

osd:
  - data:
      type: directory
      location: /var/lib/openstack-helm/ceph/osd/osd-one
    journal:
      type: directory
      location: /var/lib/openstack-helm/ceph/osd/journal-one

Host Profiles

Host profiles in Airship are tightly coupled with the hardware profiles. That means every disk or interface which is described in host profiles should have a corresponding reference to the hardware profile which is being used.

Airship always identifies every NIC or disk by its PCI or SCSI address and that means that the interfaces and the disks that are defined in host and hardware profiles should have the correct PCI and SCSI addresses objectively.

Let's give an example by following the host profile of Airsloop site.

In this Host Profile is defined that the slave interface that will be used for the pxe boot will be the pxe_nic01. That means a corresponding entry should exist in this Hardware Profile which it does. So when drydock and maas try to deploy the node it will identify the interface by the PCI address that is written in the Hardware profile.

A simple way to find out which PCI or SCSI address corresponds to which NIC or Disk is to use the lshw command. More information about that command can be found Here.

Extend Cluster

This section describes what changes need to be made to the existing manifests of Airsloop for the addition of an extra compute node to the cluster.

First and foremost the user should go to the nodes.yaml file and add an extra section for the new compute node.

The next step is to add a similar section as the existing airsloop-compute-1 section to the pki-catalog.yaml. This is essential for the correct generation of certificates and the correct communication between the nodes in the cluster.

Also every time the user adds an extra compute node to the cluster then the number of OSDs that are managed by this manifest Ceph-client should be increased by one.

Last step is to regenerate the certificates which correspond to this certificates.yaml file so the changes in the pki-catalog.yaml file takes place. This can be done through the promenade CLI.

Getting Started

Update Site Manifests.

Carefully review site manifests (site/airsloop) and update the configuration to match the hardware, networking setup and other specifics of the lab.

See more details at Site Authoring and Deployment Guide.

Note

Many manifest files (YAMLs) contain documentation in comments that instruct what changes are required for specific sections.

  1. Build Site Documents
tools/airship pegleg site -r /target collect airsloop -s collect

mkdir certs
tools/airship promenade generate-certs -o /target/certs /target/collect/*.yaml

mkdir bundle
tools/airship promenade build-all -o /target/bundle /target/collect/*.yaml /target/certs/*.yaml

See more details at Building Site documents, use site airsloop.

  1. Deploy Genesis

Deploy the Genesis node, see more details at Genesis node.

Genesis is the first node in the cluster and serves as a control node. In Airsloop configuration Genesis is the only control node (airsloop-control-1).

Airsloop is using non-bonded network interfaces:

auto lo
iface lo inet loopback

auto eno1
iface eno1 inet static
  address 10.22.70.21/24

auto enp67s0f0
iface enp67s0f0 inet manual

auto enp67s0f0.71
iface enp67s0f0.71 inet static
  address 10.22.71.21/24
  gateway 10.22.71.1
  dns-nameservers 8.8.8.8 8.8.4.4
  vlan-raw-device enp67s0f0
  vlan_id 71

auto enp67s0f0.72
iface enp67s0f0.72 inet static
  address 10.22.72.21/24
  vlan-raw-device enp67s0f0
  vlan_id 72

auto enp67s0f0.73
iface enp67s0f0.73 inet static
  address 10.22.73.21/24
  vlan-raw-device enp67s0f0
  vlan_id 73

auto enp67s0f0.74
iface enp67s0f0.74 inet static
  address 10.22.74.21/24
  vlan-raw-device enp67s0f0
  vlan_id 74

Execute Genesis bootstrap script on the Genesis server.

sudo ./genesis.sh
  1. Deploy Site
tools/airship shipyard create configdocs design --directory=/target/collect
tools/airship shipyard commit configdocs

tools/airship shipyard create action deploy_site

tools/shipyard get actions

See more details at Deploy Site with Shipyard.

Deploying Behind a Proxy

The following documents show the main differences you need to make in order to have airsloop run behind a proxy.

Note

The "-" sign refers to a line that needs to be omitted (replaced), and the "+" sign refers to a line replacing the omitted line, or simply a line that needs to be added to your yaml.

Under site/airsloop/software/charts/osh/openstack-glance/ create a glance.yaml file as follows:

---
schema: armada/Chart/v1
metadata:
  schema: metadata/Document/v1
  replacement: true
  name: glance
  layeringDefinition:
    abstract: false
    layer: site
    parentSelector:
      name: glance-type
    actions:
      - method: merge
        path: .
  storagePolicy: cleartext
data:
  test:
    enabled: false
...

Under site/airsloop/software/config/ create a versions.yaml file in the following format:

---
data:
  charts:
    kubernetes:
      apiserver:
        proxy_server: proxy.example.com:8080
      apiserver-htk:
        proxy_server: proxy.example.com:8080
      calico:
        calico:
          proxy_server: proxy.example.com:8080
        calico-htk:
          proxy_server: proxy.example.com:8080
        etcd:
          proxy_server: proxy.example.com:8080
        etcd-htk:
          proxy_server: proxy.example.com:8080
      controller-manager:
        proxy_server: proxy.example.com:8080
      controller-manager-htk:
        proxy_server: proxy.example.com:8080
      coredns:
        proxy_server: proxy.example.com:8080
      coredns-htk:
        proxy_server: proxy.example.com:8080
      etcd:
        proxy_server: proxy.example.com:8080
      etcd-htk:
        proxy_server: proxy.example.com:8080
      haproxy:
        proxy_server: proxy.example.com:8080
      haproxy-htk:
        proxy_server: proxy.example.com:8080
      ingress:
        proxy_server: proxy.example.com:8080
      ingress-htk:
        proxy_server: proxy.example.com:8080
      proxy:
        proxy_server: proxy.example.com:8080
      proxy-htk:
        proxy_server: proxy.example.com:8080
      scheduler:
        proxy_server: proxy.example.com:8080
      scheduler-htk:
        proxy_server: proxy.example.com:8080
    osh:
      barbican:
        proxy_server: proxy.example.com:8080
      cinder:
        proxy_server: proxy.example.com:8080
      cinder-htk:
        proxy_server: proxy.example.com:8080
      glance:
        proxy_server: proxy.example.com:8080
      glance-htk:
        proxy_server: proxy.example.com:8080
      heat:
        proxy_server: proxy.example.com:8080
      heat-htk:
        proxy_server: proxy.example.com:8080
      helm_toolkit:
        proxy_server: proxy.example.com:8080
      horizon:
        proxy_server: proxy.example.com:8080
      horizon-htk:
        proxy_server: proxy.example.com:8080
      ingress:
        proxy_server: proxy.example.com:8080
      ingress-htk:
        proxy_server: proxy.example.com:8080
      keystone:
        proxy_server: proxy.example.com:8080
      keystone-htk:
        proxy_server: proxy.example.com:8080
      libvirt:
        proxy_server: proxy.example.com:8080
      libvirt-htk:
        proxy_server: proxy.example.com:8080
      mariadb:
        proxy_server: proxy.example.com:8080
      mariadb-htk:
        proxy_server: proxy.example.com:8080
      memcached:
        proxy_server: proxy.example.com:8080
      memcached-htk:
        proxy_server: proxy.example.com:8080
      neutron:
        proxy_server: proxy.example.com:8080
      neutron-htk:
        proxy_server: proxy.example.com:8080
      nova:
        proxy_server: proxy.example.com:8080
      nova-htk:
        proxy_server: proxy.example.com:8080
      openvswitch:
        proxy_server: proxy.example.com:8080
      openvswitch-htk:
        proxy_server: proxy.example.com:8080
      rabbitmq:
        proxy_server: proxy.example.com:8080
      rabbitmq-htk:
        proxy_server: proxy.example.com:8080
      tempest:
        proxy_server: proxy.example.com:8080
      tempest-htk:
        proxy_server: proxy.example.com:8080
    osh_infra:
      elasticsearch:
        proxy_server: proxy.example.com:8080
      fluentbit:
        proxy_server: proxy.example.com:8080
      fluentd:
        proxy_server: proxy.example.com:8080
      grafana:
        proxy_server: proxy.example.com:8080
      helm_toolkit:
        proxy_server: proxy.example.com:8080
      kibana:
        proxy_server: proxy.example.com:8080
      nagios:
        proxy_server: proxy.example.com:8080
      nfs_provisioner:
        proxy_server: proxy.example.com:8080
      podsecuritypolicy:
        proxy_server: proxy.example.com:8080
      prometheus:
        proxy_server: proxy.example.com:8080
      prometheus_alertmanager:
        proxy_server: proxy.example.com:8080
      prometheus_kube_state_metrics:
        proxy_server: proxy.example.com:8080
      prometheus_node_exporter:
        proxy_server: proxy.example.com:8080
      prometheus_openstack_exporter:
        proxy_server: proxy.example.com:8080
      prometheus_process_exporter:
        proxy_server: proxy.example.com:8080
    ucp:
      armada:
        proxy_server: proxy.example.com:8080
      armada-htk:
        proxy_server: proxy.example.com:8080
      barbican:
        proxy_server: proxy.example.com:8080
      barbican-htk:
        proxy_server: proxy.example.com:8080
      ceph-client:
        proxy_server: proxy.example.com:8080
      ceph-htk:
        proxy_server: proxy.example.com:8080
      ceph-mon:
        proxy_server: proxy.example.com:8080
      ceph-osd:
        proxy_server: proxy.example.com:8080
      ceph-provisioners:
        proxy_server: proxy.example.com:8080
      ceph-rgw:
        proxy_server: proxy.example.com:8080
      deckhand:
        proxy_server: proxy.example.com:8080
      deckhand-htk:
        proxy_server: proxy.example.com:8080
      divingbell:
        proxy_server: proxy.example.com:8080
      divingbell-htk:
        proxy_server: proxy.example.com:8080
      drydock:
        proxy_server: proxy.example.com:8080
      drydock-htk:
        proxy_server: proxy.example.com:8080
      ingress:
        proxy_server: proxy.example.com:8080
      ingress-htk:
        proxy_server: proxy.example.com:8080
      keystone:
        proxy_server: proxy.example.com:8080
      keystone-htk:
        proxy_server: proxy.example.com:8080
      maas:
        proxy_server: proxy.example.com:8080
      maas-htk:
        proxy_server: proxy.example.com:8080
      mariadb:
        proxy_server: proxy.example.com:8080
      mariadb-htk:
        proxy_server: proxy.example.com:8080
      memcached:
        proxy_server: proxy.example.com:8080
      memcached-htk:
        proxy_server: proxy.example.com:8080
      postgresql:
        proxy_server: proxy.example.com:8080
      postgresql-htk:
        proxy_server: proxy.example.com:8080
      promenade:
        proxy_server: proxy.example.com:8080
      promenade-htk:
        proxy_server: proxy.example.com:8080
      rabbitmq:
        proxy_server: proxy.example.com:8080
      rabbitmq-htk:
        proxy_server: proxy.example.com:8080
      shipyard:
        proxy_server: proxy.example.com:8080
      shipyard-htk:
        proxy_server: proxy.example.com:8080
      tenant-ceph-client:
        proxy_server: proxy.example.com:8080
      tenant-ceph-htk:
        proxy_server: proxy.example.com:8080
      tenant-ceph-mon:
        proxy_server: proxy.example.com:8080
      tenant-ceph-osd:
        proxy_server: proxy.example.com:8080
      tenant-ceph-provisioners:
        proxy_server: proxy.example.com:8080
      tenant-ceph-rgw:
        proxy_server: proxy.example.com:8080
      tiller:
        proxy_server: proxy.example.com:8080
      tiller-htk:
        proxy_server: proxy.example.com:8080
metadata:
  name: software-versions
  replacement: true
  layeringDefinition:
    abstract: false
    layer: site
    parentSelector:
      name: software-versions-global
    actions:
      - method: merge
        path: .
  storagePolicy: cleartext
  schema: metadata/Document/v1
schema: pegleg/SoftwareVersions/v1
...

Update site/airsloop/networks/common-addresses.yaml to add the proxy information as follows:

# settings are correct and reachable in your environment; otherwise update
# them with the correct values for your environment.
proxy:
-    http: ""
-    https: ""
-    no_proxy: []
+    http: "proxy.example.com:8080"
+    https: "proxy.example.com:8080"
+    no_proxy:
+      - 127.0.0.1

Under site/airsloop/software/charts/ucp/ create the file maas.yaml with the following format:

---
# This file defines site-specific deviations for MaaS.
schema: armada/Chart/v1
metadata:
  schema: metadata/Document/v1
  replacement: true
  name: ucp-maas
  layeringDefinition:
    abstract: false
    layer: site
    parentSelector:
      name: ucp-maas-type
    actions:
      - method: merge
        path: .
  storagePolicy: cleartext
data:
    values:
      conf:
        maas:
          proxy:
            proxy_enabled: true
            peer_proxy_enabled: true
            proxy_server: 'http://proxy.example.com:8080'
...

Under site/airsloop/software/charts/ucp/ create a promenade.yaml file in the following format:

---
# This file defines site-specific deviations for Promenade.
schema: armada/Chart/v1
metadata:
  schema: metadata/Document/v1
  replacement: true
  name: ucp-promenade
  layeringDefinition:
    abstract: false
    layer: site
    parentSelector:
      name: ucp-promenade-type
    actions:
      - method: merge
        path: .
  storagePolicy: cleartext
data:
  values:
    pod:
      env:
        promenade_api:
          - name: http_proxy
            value: http://proxy.example.com:8080
          - name: https_proxy
            value: http://proxy.example.com:8080
          - name: no_proxy
            value: "127.0.0.1,localhost,kubernetes,kubernetes.default,kubernetes.default.svc,kubernetes.default.svc.cluster.local,.cluster.local"
          - name: HTTP_PROXY
            value: http://proxy.example.com:8080
          - name: HTTP_PROXY
            value: http://proxy.example.com:8080
          - name: HTTPS_PROXY
            value: http://proxy.example.com:8080
          - name: NO_PROXY
            value: "127.0.0.1,localhost,kubernetes,kubernetes.default,kubernetes.default.svc,kubernetes.default.svc.cluster.local,.cluster.local"
...