github/kubernetes
4
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Merge pull request #7998 from doublerr/rackspace_fixes

Browse Source 
Rackspace fixes
This commit is contained in:
Eric Tune 2015-05-19 09:59:54 -07:00
commit dc81fe1f12
7 changed files with 285 additions and 137 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

36
cluster/rackspace/authorization.sh Normal file
View File

@ -0,0 +1,36 @@
#!/bin/bash
# Copyright 2015 The Kubernetes Authors All rights reserved.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
# Create generic token following GCE standard
create_token() {
echo $(cat /dev/urandom | base64 | tr -d "=+/" | dd bs=32 count=1 2> /dev/null)
}
get_token_from_csv() {
KUBE_BEARER_TOKEN=$(awk -F, '/admin/ {print $1}' ${KUBE_TEMP}/${1}_tokens.csv)
KUBELET_TOKEN=$(awk -F, '/kubelet/ {print $1}' ${KUBE_TEMP}/${1}_tokens.csv)
KUBE_PROXY_TOKEN=$(awk -F, '/kube_proxy/ {print $1}' ${KUBE_TEMP}/${1}_tokens.csv)
}
generate_admin_token() {
echo "$(create_token),admin,admin" >> ${KUBE_TEMP}/${1}_tokens.csv
}
# Creates a csv file each time called (i.e one per kubelet).
generate_kubelet_tokens() {
echo "$(create_token),kubelet,kubelet" > ${KUBE_TEMP}/${1}_tokens.csv
echo "$(create_token),kube_proxy,kube_proxy" >> ${KUBE_TEMP}/${1}_tokens.csv
}

77
cluster/rackspace/cloud-config/master-cloud-config.yaml
View File

@ -23,12 +23,19 @@ write_files:
permissions: 0755
content: |
#!/bin/bash
# This temp URL is only good for the length of time specified at cluster creation time.
# Afterward, it will result in a 403.
OBJECT_URL="CLOUD_FILES_URL"
if [ ! -s /opt/kubernetes.tar.gz ]
then
echo "Downloading release ($OBJECT_URL)"
wget "${OBJECT_URL}" -O /opt/kubernetes.tar.gz
echo "Unpacking release"
rm -rf /opt/kubernetes || false
tar xzf /opt/kubernetes.tar.gz -C /opt/
else
echo "kubernetes release found. Skipping download."
fi
- path: /opt/.kubernetes_auth
permissions: 0600
content: |
@ -65,7 +72,7 @@ coreos:
Type=oneshot
RemainAfterExit=yes
ExecStart=/usr/bin/bash /opt/bin/download-release.sh
- name: master-apiserver.service
- name: kube-apiserver.service
command: start
content: |
[Unit]
@ -77,10 +84,33 @@ coreos:
Requires=download-release.service
[Service]
ExecStartPre=/usr/bin/ln -sf /opt/kubernetes/server/bin/kube-apiserver /opt/bin/kube-apiserver
ExecStart=/opt/bin/kube-apiserver --address=127.0.0.1 --port=8080 --etcd_servers=http://127.0.0.1:4001 --portal_net=PORTAL_NET --logtostderr=true --cloud_provider=rackspace --cloud_config=/etc/cloud.conf --v=2
ExecStartPre=/usr/bin/mkdir -p /var/lib/kube-apiserver
ExecStartPre=/usr/bin/cp /media/configdrive/openstack/content/0000 /var/lib/kube-apiserver/known_tokens.csv
ExecStart=/opt/bin/kube-apiserver \
--address=127.0.0.1 \
--cloud_provider=rackspace \
--cloud_config=/etc/cloud.conf \
--etcd_servers=http://127.0.0.1:4001 \
--logtostderr=true \
--port=8080 \
--portal_net=PORTAL_NET \
--token-auth-file=/var/lib/kube-apiserver/known_tokens.csv \
--v=2
Restart=always
RestartSec=2
- name: master-controller-manager.service
RestartSec=5
- name: apiserver-advertiser.service
command: start
content: |
[Unit]
Description=Kubernetes Apiserver Advertiser
After=etcd.service
Requires=etcd.service
After=master-apiserver.service
[Service]
ExecStart=/bin/sh -c 'etcdctl set /corekube/apiservers/$public_ipv4 $public_ipv4'
Restart=always
RestartSec=120
- name: kube-controller-manager.service
command: start
content: |
[Unit]
@ -88,14 +118,19 @@ coreos:
Documentation=https://github.com/GoogleCloudPlatform/kubernetes
After=network-online.target
Requires=network-online.target
After=master-apiserver.service
Requires=master-apiserver.service
After=kube-apiserver.service
Requires=kube-apiserver.service
[Service]
ExecStartPre=/usr/bin/ln -sf /opt/kubernetes/server/bin/kube-controller-manager /opt/bin/kube-controller-manager
ExecStart=/opt/bin/kube-controller-manager --master=127.0.0.1:8080 --logtostderr=true --cloud_provider=rackspace --cloud_config=/etc/cloud.conf --v=2
ExecStart=/opt/bin/kube-controller-manager \
--cloud_provider=rackspace \
--cloud_config=/etc/cloud.conf \
--logtostderr=true \
--master=127.0.0.1:8080 \
--v=2
Restart=always
RestartSec=2
- name: master-scheduler.service
RestartSec=5
- name: kube-scheduler.service
command: start
content: |
[Unit]
@ -103,27 +138,33 @@ coreos:
Documentation=https://github.com/GoogleCloudPlatform/kubernetes
After=network-online.target
Requires=network-online.target
After=master-apiserver.service
Requires=master-apiserver.service
After=kube-apiserver.service
Requires=kube-apiserver.service
[Service]
ExecStartPre=/usr/bin/ln -sf /opt/kubernetes/server/bin/kube-scheduler /opt/bin/kube-scheduler
ExecStart=/opt/bin/kube-scheduler --master=127.0.0.1:8080 --logtostderr=true
ExecStart=/opt/bin/kube-scheduler \
--logtostderr=true \
--master=127.0.0.1:8080
Restart=always
RestartSec=10
- name: master-register.service
RestartSec=5
- name: kube-register.service
command: start
content: |
[Unit]
Description=Kubernetes Registration Service
Documentation=https://github.com/kelseyhightower/kube-register
Requires=kube-apiserver.service
After=kube-apiserver.service
Requires=fleet.service
After=fleet.service
[Service]
ExecStartPre=/usr/bin/wget -N -P /opt/bin https://storage.googleapis.com/kubernetes/kube-register
ExecStartPre=/usr/bin/curl -L -o /opt/bin/kube-register -z /opt/bin/kube-register https://github.com/kelseyhightower/kube-register/releases/download/v0.0.3/kube-register-0.0.3-linux-amd64
ExecStartPre=/usr/bin/chmod +x /opt/bin/kube-register
ExecStart=/opt/bin/kube-register \
--metadata=kubernetes_role=minion \
--api-endpoint=http://127.0.0.1:8080 \
--fleet-endpoint=unix:///var/run/fleet.sock \
--api-endpoint=http://127.0.0.1:8080
--healthz-port=10248 \
--metadata=kubernetes_role=minion
Restart=always
RestartSec=10
#Running nginx service with --net="host" is a necessary evil until running all k8s services in docker.

198
cluster/rackspace/cloud-config/minion-cloud-config.yaml
View File

@ -1,27 +1,62 @@
#cloud-config
write_files:
- path: /opt/bin/kube-net-update.sh
- path: /opt/bin/regen-apiserver-list.sh
permissions: 0755
content: |
#!/bin/sh
set -x -e
nh=${ETCD_WATCH_KEY##*/}
net=$ETCD_WATCH_VALUE
case $ETCD_WATCH_ACTION in
set) ip route replace $net via $nh dev eth2 metric 900 ;;
expire) ip route del $net via $nh metric 900 ;;
esac
m=$(echo $(etcdctl ls --recursive /corekube/apiservers | cut -d/ -f4 | sort) | tr ' ' ,)
mkdir -p /run/kubelet
echo "APISERVER_IPS=$m" > /run/kubelet/apiservers.env
echo "FIRST_APISERVER_URL=https://${m%%\,*}:6443" >> /run/kubelet/apiservers.env
- path: /opt/bin/download-release.sh
permissions: 0755
content: |
#!/bin/bash
# This temp URL is only good for the length of time specified at cluster creation time.
# Afterward, it will result in a 403.
OBJECT_URL="CLOUD_FILES_URL"
if [ ! -s /opt/kubernetes.tar.gz ]
then
echo "Downloading release ($OBJECT_URL)"
wget "${OBJECT_URL}" -O /opt/kubernetes.tar.gz
echo "Unpacking release"
rm -rf /opt/kubernetes || false
tar xzf /opt/kubernetes.tar.gz -C /opt/
else
echo "kubernetes release found. Skipping download."
fi
- path: /run/setup-auth.sh
permissions: 0755
content: |
#!/bin/bash -e
set -x
/usr/bin/mkdir -p /var/lib/kubelet
/bin/echo "{\"BearerToken\": \"KUBELET_TOKEN\", \"Insecure\": true }" > /var/lib/kubelet/kubernetes_auth
- path: /run/config-kube-proxy.sh
permissions: 0755
content: |
#!/bin/bash -e
set -x
/usr/bin/mkdir -p /var/lib/kube-proxy
cat > /var/lib/kube-proxy/kubeconfig << EOF
apiVersion: v1
kind: Config
users:
- name: kube-proxy
user:
token: KUBE_PROXY_TOKEN
clusters:
- name: local
cluster:
insecure-skip-tls-verify: true
contexts:
- context:
cluster: local
user: kube-proxy
name: service-account-context
current-context: service-account-context
EOF
coreos:
etcd:
@ -43,6 +78,31 @@ coreos:
command: start
- name: fleet.service
command: start
- name: flanneld.service
command: start
drop-ins:
- name: 50-network-config.conf
content: |
[Service]
ExecStartPre=/usr/bin/etcdctl set /coreos.com/network/config '{ "Network": "KUBE_NETWORK", "Backend": {"Type": "host-gw"}}'
ExecStart=
ExecStart=/usr/libexec/sdnotify-proxy /run/flannel/sd.sock \
/usr/bin/docker run --net=host --privileged=true --rm \
--volume=/run/flannel:/run/flannel \
--env=NOTIFY_SOCKET=/run/flannel/sd.sock \
--env-file=/run/flannel/options.env \
--volume=${ETCD_SSL_DIR}:/etc/ssl/etcd:ro \
quay.io/coreos/flannel:${FLANNEL_VER} /opt/bin/flanneld -etcd-endpoints http://127.0.0.1:4001 --ip-masq=true --iface=eth2
- name: docker.service
command: start
drop-ins:
- name: 51-docker-mirror.conf
content: |
[Unit]
# making sure that flanneld finished startup, otherwise containers
# won't land in flannel's network...
Requires=flanneld.service
After=flanneld.service
- name: download-release.service
command: start
content: |
@ -54,7 +114,7 @@ coreos:
Type=oneshot
RemainAfterExit=yes
ExecStart=/usr/bin/bash /opt/bin/download-release.sh
- name: minion-kubelet.service
- name: kubelet.service
command: start
content: |
[Unit]
@ -67,11 +127,21 @@ coreos:
After=download-release.service
Requires=download-release.service
[Service]
EnvironmentFile=/run/kubelet/apiservers.env
ExecStartPre=/run/setup-auth.sh
ExecStartPre=/usr/bin/ln -sf /opt/kubernetes/server/bin/kubelet /opt/bin/kubelet
ExecStart=/opt/bin/kubelet --address=$private_ipv4 --hostname_override=$private_ipv4 --etcd_servers=http://127.0.0.1:4001 --logtostderr=true --config=/opt/kubernetes-manifests --cluster_dns=DNS_SERVER_IP --cluster_domain=DNS_DOMAIN
ExecStart=/opt/bin/kubelet \
--address=$private_ipv4 \
--api_servers=${FIRST_APISERVER_URL} \
--cluster_dns=DNS_SERVER_IP \
--cluster_domain=DNS_DOMAIN \
--healthz-bind-address=$private_ipv4 \
--hostname_override=$private_ipv4 \
--logtostderr=true \
--v=2
Restart=always
RestartSec=2
- name: minion-proxy.service
RestartSec=5
- name: kube-proxy.service
command: start
content: |
[Unit]
@ -84,46 +154,50 @@ coreos:
After=download-release.service
Requires=download-release.service
[Service]
EnvironmentFile=/run/kubelet/apiservers.env
ExecStartPre=/run/config-kube-proxy.sh
ExecStartPre=/usr/bin/ln -sf /opt/kubernetes/server/bin/kube-proxy /opt/bin/kube-proxy
ExecStart=/opt/bin/kube-proxy --bind_address=$private_ipv4 --etcd_servers=http://127.0.0.1:4001 --logtostderr=true --v=2
ExecStart=/opt/bin/kube-proxy \
--bind_address=$private_ipv4 \
--kubeconfig=/var/lib/kube-proxy/kubeconfig \
--logtostderr=true \
--master=${FIRST_APISERVER_URL}
Restart=always
RestartSec=2
- name: minion-advertiser.service
RestartSec=5
- name: kubelet-sighup.path
command: start
content: |
[Path]
PathChanged=/run/kubelet/apiservers.env
- name: kubelet-sighup.service
command: start
content: |
[Service]
ExecStart=/usr/bin/pkill -SIGHUP -f kubelet
- name: kube-proxy-sighup.path
command: start
content: |
[Path]
PathChanged=/run/kubelet/apiservers.env
- name: kube-proxy-sighup.service
command: start
content: |
[Service]
ExecStart=/usr/bin/pkill -SIGHUP -f kube-proxy
- name: apiserver-finder.service
command: start
content: |
[Unit]
Description=Kubernetes Minion Advertiser
Description=Kubernetes Apiserver finder
After=network-online.target
Requires=network-online.target
After=etcd.service
Requires=etcd.service
After=minion-kubelet.service
[Service]
ExecStart=/bin/sh -c 'while :; do etcdctl set /corekube/minions/$private_ipv4 $private_ipv4 --ttl 300; sleep 120; done'
ExecStartPre=/opt/bin/regen-apiserver-list.sh
ExecStart=/usr/bin/etcdctl exec-watch --recursive /corekube/apiservers -- /opt/bin/regen-apiserver-list.sh
Restart=always
RestartSec=120
- name: net-advertiser.service
command: start
content: |
[Unit]
Description=Kubernetes Network Advertiser
After=etcd.service
Requires=etcd.service
After=minion-kubelet.service
[Service]
ExecStart=/bin/sh -c 'eth2_ip=$$(ip -o -f inet a show dev eth2 | sed "s/.* inet \([0-9.]\+\).*/\1/"); while :; do etcdctl set /corekube/net/$$eth2_ip 10.240.INDEX.0/24 --ttl 300; sleep 120; done'
Restart=always
RestartSec=120
- name: net-router.service
command: start
content: |
[Unit]
Description=Kubernetes Network Router
After=etcd.service
Requires=etcd.service
After=minion-kubelet.service
[Service]
ExecStart=/usr/bin/etcdctl exec-watch --recursive /corekube/net -- /opt/bin/kube-net-update.sh
Restart=always
RestartSec=120
RestartSec=30
- name: cbr0.netdev
command: start
content: |
@ -149,41 +223,3 @@ coreos:
ExecStart=/usr/sbin/iptables -t nat -A POSTROUTING -o eth1 -s 10.240.INDEX.0/24 -j MASQUERADE
RemainAfterExit=yes
Type=oneshot
- name: docker.service
command: start
content: |
[Unit]
After=network.target
Description=Docker Application Container Engine
Documentation=http://docs.docker.io
[Service]
ExecStartPre=/bin/mount --make-rprivate /
ExecStart=/usr/bin/docker -d -H fd:// -b cbr0 --iptables=false
Restart=always
RestartSec=30
[Install]
WantedBy=multi-user.target
- name: format-data.service
command: start
content: |
[Unit]
Description=Formats data drive
[Service]
Type=oneshot
RemainAfterExit=yes
ExecStart=/usr/sbin/wipefs -f /dev/xvde1
ExecStart=/usr/sbin/mkfs.btrfs -f /dev/xvde1
- name: var-lib-docker-volumes.mount
command: start
content: |
[Unit]
Description=Mount data drive to /var/lib/docker/volumes
Requires=format-data.service
After=format-data.service
Before=docker.service
[Mount]
What=/dev/xvde1
Where=/var/lib/docker/volumes
Type=btrfs

4
cluster/rackspace/config-default.sh
View File

@ -19,7 +19,7 @@
# KUBE_IMAGE, KUBE_MASTER_FLAVOR, KUBE_MINION_FLAVOR, NUM_MINIONS, NOVA_NETWORK and SSH_KEY_NAME
# Shared
KUBE_IMAGE="${KUBE_IMAGE-b63e1435-a46f-4726-b984-e3f15ae92753}" # CoreOS(Beta)
KUBE_IMAGE="${KUBE_IMAGE-2c210e44-5149-4ae3-83d6-f855a4d28490}" # CoreOS(Beta)
SSH_KEY_NAME="${SSH_KEY_NAME-id_kubernetes}"
NOVA_NETWORK_LABEL="kubernetes-pool-net"
NOVA_NETWORK_CIDR="${NOVA_NETWORK-192.168.0.0/24}"
@ -35,7 +35,7 @@ KUBE_MINION_FLAVOR="${KUBE_MINION_FLAVOR-performance1-2}"
RAX_NUM_MINIONS="${RAX_NUM_MINIONS-4}"
MINION_TAG="tags=${INSTANCE_PREFIX}-minion"
MINION_NAMES=($(eval echo ${INSTANCE_PREFIX}-minion-{1..${RAX_NUM_MINIONS}}))
KUBE_NETWORK=($(eval echo "10.240.{1..${RAX_NUM_MINIONS}}.0/24"))
KUBE_NETWORK="10.240.0.0/16"
PORTAL_NET="10.0.0.0/16"
# Optional: Install node monitoring.

45
cluster/rackspace/util.sh
View File

@ -21,6 +21,7 @@
KUBE_ROOT=$(dirname "${BASH_SOURCE}")/../..
source $(dirname ${BASH_SOURCE})/${KUBE_CONFIG_FILE-"config-default.sh"}
source "${KUBE_ROOT}/cluster/common.sh"
source "${KUBE_ROOT}/cluster/rackspace/authorization.sh"
verify-prereqs() {
# Make sure that prerequisites are installed.
@ -60,7 +61,7 @@ get-password() {
get-kubeconfig-basicauth
if [[ -z "${KUBE_USER}" || -z "${KUBE_PASSWORD}" ]]; then
KUBE_USER=admin
KUBE_PASSWORD=$(python -c 'import string,random; print "".join(random.SystemRandom().choice(string.ascii_letters + string.digits) for _ in range(16))')
KUBE_PASSWORD=$(python2.7 -c 'import string,random; print "".join(random.SystemRandom().choice(string.ascii_letters + string.digits) for _ in range(16))')
fi
}
@ -106,7 +107,8 @@ find-object-url() {
KUBE_TAR=${CLOUDFILES_CONTAINER}/${CONTAINER_PREFIX}/kubernetes-server-linux-amd64.tar.gz
RELEASE_TMP_URL=$(swiftly -A ${OS_AUTH_URL} -U ${OS_USERNAME} -K ${OS_PASSWORD} tempurl GET ${KUBE_TAR})
# Create temp URL good for 24 hours
RELEASE_TMP_URL=$(swiftly -A ${OS_AUTH_URL} -U ${OS_USERNAME} -K ${OS_PASSWORD} tempurl GET ${KUBE_TAR} 86400 )
echo "cluster/rackspace/util.sh: Object temp URL:"
echo -e "\t${RELEASE_TMP_URL}"
@ -132,6 +134,25 @@ copy_dev_tarballs() {
echo "Release pushed."
}
prep_known_tokens() {
for (( i=0; i<${#MINION_NAMES[@]}; i++)); do
generate_kubelet_tokens ${MINION_NAMES[i]}
cat ${KUBE_TEMP}/${MINION_NAMES[i]}_tokens.csv >> ${KUBE_TEMP}/known_tokens.csv
done
# Generate tokens for other "service accounts". Append to known_tokens.
#
# NB: If this list ever changes, this script actually has to
# change to detect the existence of this file, kill any deleted
# old tokens and add any new tokens (to handle the upgrade case).
local -r service_accounts=("system:scheduler" "system:controller_manager" "system:logging" "system:monitoring" "system:dns")
for account in "${service_accounts[@]}"; do
echo "$(create_token),${account},${account}" >> ${KUBE_TEMP}/known_tokens.csv
done
generate_admin_token
}
rax-boot-master() {
DISCOVERY_URL=$(curl https://discovery.etcd.io/new)
@ -159,6 +180,7 @@ rax-boot-master() {
--meta ${MASTER_TAG} \
--meta ETCD=${DISCOVERY_ID} \
--user-data ${KUBE_TEMP}/master-cloud-config.yaml \
--file /var/lib/kube-apiserver/known_tokens.csv=${KUBE_TEMP}/known_tokens.csv \
--config-drive true \
--nic net-id=${NETWORK_UUID} \
${MASTER_NAME}"
@ -175,15 +197,20 @@ rax-boot-minions() {
for (( i=0; i<${#MINION_NAMES[@]}; i++)); do
get_tokens_from_csv ${MINION_NAMES[i]}
sed -e "s|DISCOVERY_ID|${DISCOVERY_ID}|" \
-e "s|INDEX|$((i + 1))|g" \
-e "s|CLOUD_FILES_URL|${RELEASE_TMP_URL//&/\\&}|" \
-e "s|ENABLE_NODE_MONITORING|${ENABLE_NODE_MONITORING:-false}|" \
-e "s|ENABLE_NODE_LOGGING|${ENABLE_NODE_LOGGING:-false}|" \
-e "s|LOGGING_DESTINATION|${LOGGING_DESTINATION:-}|" \
-e "s|ENABLE_CLUSTER_DNS|${ENABLE_CLUSTER_DNS:-false}|" \
-e "s|DNS_SERVER_IP|${DNS_SERVER_IP:-}|" \
-e "s|DNS_DOMAIN|${DNS_DOMAIN:-}|" \
-e "s|ENABLE_CLUSTER_DNS|${ENABLE_CLUSTER_DNS:-false}|" \
-e "s|ENABLE_NODE_MONITORING|${ENABLE_NODE_MONITORING:-false}|" \
-e "s|ENABLE_NODE_LOGGING|${ENABLE_NODE_LOGGING:-false}|" \
-e "s|INDEX|$((i + 1))|g" \
-e "s|KUBELET_TOKEN|${KUBELET_TOKEN}|" \
-e "s|KUBE_NETWORK|${KUBE_NETWORK}|" \
-e "s|KUBE_PROXY_TOKEN|${KUBE_PROXY_TOKEN}|" \
-e "s|LOGGING_DESTINATION|${LOGGING_DESTINATION:-}|" \
$(dirname $0)/rackspace/cloud-config/minion-cloud-config.yaml > $KUBE_TEMP/minion-cloud-config-$(($i + 1)).yaml
@ -276,7 +303,7 @@ kube-up() {
trap "rm -rf ${KUBE_TEMP}" EXIT
get-password
python $(dirname $0)/../third_party/htpasswd/htpasswd.py -b -c ${KUBE_TEMP}/htpasswd $KUBE_USER $KUBE_PASSWORD
python2.7 $(dirname $0)/../third_party/htpasswd/htpasswd.py -b -c ${KUBE_TEMP}/htpasswd $KUBE_USER $KUBE_PASSWORD
HTPASSWD=$(cat ${KUBE_TEMP}/htpasswd)
rax-nova-network
@ -286,6 +313,8 @@ kube-up() {
rax-ssh-key
echo "cluster/rackspace/util.sh: Starting Cloud Servers"
prep_known_tokens
rax-boot-master
rax-boot-minions

1
docs/getting-started-guides/README.md
View File

@ -39,6 +39,7 @@ Docker Multi Node | Flannel| N/A | local | [docs](docker-multino
Local | | | _none_ | [docs](../../docs/getting-started-guides/locally.md) | Community (@preillyme) |
libvirt/KVM | CoreOS | CoreOS | libvirt/KVM | [docs](../../docs/getting-started-guides/libvirt-coreos.md) | Community (@lhuard1A) |
oVirt | | | | [docs](../../docs/getting-started-guides/ovirt.md) | Community (@simon3z) |
Rackspace | CoreOS | CoreOS | flannel | [docs](../../docs/getting-started-guides/rackspace.md) | Community (@doublerr) | use k8s version 0.16.2
*Note*: The above table is ordered by version test/used in notes followed by support level.

35
docs/getting-started-guides/rackspace.md
View File

@ -1,25 +1,27 @@
# Status: Out Of Date
** Rackspace support is out of date. Please check back later **
# Rackspace
In general, the dev-build-and-up.sh workflow for Rackspace is the similar to GCE. The specific implementation is different due to the use of CoreOS, Rackspace Cloud Files and network design.
* Supported Version: v0.16.2
* `git checkout v0.16.2`
In general, the dev-build-and-up.sh workflow for Rackspace is the similar to GCE. The specific implementation is different due to the use of CoreOS, Rackspace Cloud Files and the overall network design.
These scripts should be used to deploy development environments for Kubernetes. If your account leverages RackConnect or non-standard networking, these scripts will most likely not work without modification.
NOTE: The rackspace scripts do NOT rely on `saltstack`.
NOTE: The rackspace scripts do NOT rely on `saltstack` and instead rely on cloud-init for configuration.
The current cluster design is inspired by:
- [corekube](https://github.com/metral/corekube/)
- [Angus Lees](https://github.com/anguslees/kube-openstack/)
## Prerequisites
1. You need to have both `nova` and `swiftly` installed. It's recommended to use a python virtualenv to install these packages into.
2. Make sure you have the appropriate environment variables set to interact with the OpenStack APIs. See [Rackspace Documentation](http://docs.rackspace.com/servers/api/v2/cs-gettingstarted/content/section_gs_install_nova.html) for more details.
1. Python2.7
2. You need to have both `nova` and `swiftly` installed. It's recommended to use a python virtualenv to install these packages into.
3. Make sure you have the appropriate environment variables set to interact with the OpenStack APIs. See [Rackspace Documentation](http://docs.rackspace.com/servers/api/v2/cs-gettingstarted/content/section_gs_install_nova.html) for more details.
##Provider: Rackspace
- To use Rackspace as the provider, set the KUBERNETES_PROVIDER ENV variable:
`export KUBERNETES_PROVIDER=rackspace` and run the `bash hack/dev-build-and-up.sh` script.
- To install the latest released version of kubernetes use `export KUBERNETES_PROVIDER=rackspace; wget -q -O - https://get.k8s.io | bash`
- To build your own released version from source use `export KUBERNETES_PROVIDER=rackspace` and run the `bash hack/dev-build-and-up.sh`
## Build
1. The kubernetes binaries will be built via the common build scripts in `build/`.
@ -28,16 +30,19 @@ The current cluster design is inspired by:
3. The built `kubernetes-server-linux-amd64.tar.gz` will be uploaded to this container and the URL will be passed to master/minions nodes when booted.
## Cluster
1. There is a specific `cluster/rackspace` directory with the scripts for the following steps:
2. A cloud network will be created and all instances will be attached to this network. We will connect the master API and minion kubelet service via this network.
3. A SSH key will be created and uploaded if needed. This key must be used to ssh into the machines since we won't capture the password.
4. A master and minions will be created via the `nova` CLI. A `cloud-config.yaml` is generated and provided as user-data with the entire configuration for the systems.
5. We then boot as many minions as defined via `$RAX_NUM_MINIONS`.
There is a specific `cluster/rackspace` directory with the scripts for the following steps:
1. A cloud network will be created and all instances will be attached to this network.
- flanneld uses this network for next hop routing. These routes allow the containers running on each node to communicate with one another on this private network.
2. A SSH key will be created and uploaded if needed. This key must be used to ssh into the machines since we won't capture the password.
3. The master server and additional nodes will be created via the `nova` CLI. A `cloud-config.yaml` is generated and provided as user-data with the entire configuration for the systems.
4. We then boot as many nodes as defined via `$RAX_NUM_MINIONS`.
## Some notes:
- The scripts expect `eth2` to be the cloud network that the containers will communicate across.
- A number of the items in `config-default.sh` are overridable via environment variables.
- For older versions please either:
* Sync back to `v0.9` with `git checkout v0.9`
* Download a [snapshot of `v0.9`](https://github.com/GoogleCloudPlatform/kubernetes/archive/v0.9.tar.gz)
* Sync back to `v0.3` with `git checkout v0.3`
* Download a [snapshot of `v0.3`](https://github.com/GoogleCloudPlatform/kubernetes/archive/v0.3.tar.gz)