- Vagrant SES build validation
- Set-up Vagrant server
- Requiered packages
- Install SUSE certificate
- Configure QEMU to run processes as root user
- Changing Libvirt default pool location
- Enable vagrant-libvirt network autostart
- Enable symbolic links in Apache DocumentRoot
- Create repo files and repo directories
- Enable and start services:
- Firewall
- SSH keys
- Setting up Docker registry cache
- Change Vagrant home directory location
- NFS Server settings
- Additional server configuration for testing SES cluster with mixed architecture
- Preparing for Build Validation tests
- Vagrant with OpenStack
- Running Build Validation tests
- Set-up Vagrant server
Using native Vagrant files to build cluster with SES and running build validaion on it.
We have already everything ready and using server jupiter.qa.suse.cz for x86_64 or thunderx7.arch.suse.de for aarch64 testing.
To login to the servers as root user use storage-automation key.
For Build Validation it's enough to trigger the test on Jenkins
or to follow the steps described in:
You'll need following to have installed on your OpenSuse server:
-
Libvirt
-
QEMU
-
NFS Server
-
Docker
-
Apache
-
vagrant
-
virt-install package
-
pdsh (if you going to trigger build validation with build_validation.sh)
zypper in vagrant docker apache2 patterns-server-kvm_servervagrant plugin install vagrant-hostsupdater
vagrant plugin install vagrant-libvirtGo to http://download.suse.de/ibs/SUSE:/CA/ and choose distro you are running.
Add it to repositories (Tumbleweed in our case) and install SUSE CA
zypper ar -f http://download.suse.de/ibs/SUSE:/CA/openSUSE_Tumbleweed/SUSE:CA.repo
zypper in -y ca-certificates-suse
/etc/libvirt/qemu.conf
user = root
group = rootIn this setup we prefer Libvirt to use default pool location on separate disk / mountpoint as is its default. If another disk is available on your server then mount it under /qemu . Ohterwise you can use Libvirt default settings unless you created separate partition for /qemu.
mkdir -p /qemu/pools/default
virsh pool-destroy default
virsh pool-undefine default
virsh pool-create-as --name default --type dir --target /qemu/pools/default
virsh pool-dumpxml default > /tmp/default
virsh pool-define --file /tmp/default
virsh pool-autostart defaultWe prefer Libvirt network vagratn-libvirt to start automatically
virsh net-autostart vagrant-libvirt/etc/apache2/default-server.conf
add FollowSymLinks into the Options
<Directory "/srv/www/htdocs">
Options FollowSymLinks
AllowOverride None
<IfModule !mod_access_compat.c>
Require all granted
</IfModule>
<IfModule mod_access_compat.c>
Order allow,deny
Allow from all
</IfModule>
</Directory>Create directories under Apache DocumentRoot:
mkdir /srv/www/htdocs/current_os
mkdir /srv/www/htdocs/current_ses
Create repo files under Apache DocumentRoot:
Our libvirt NIC IP is 192.168.122.1
ip addr show virbr0
curretn_os.repo file:
# cat /srv/www/htdocs/current_os.repo
[basesystem]
name=basesystem
type=rpm-md
baseurl=http://192.168.122.1/current_os/Module-Basesystem/
gpgcheck=0
gpgkey=http://192.168.122.1/current_os/Module-Basesystem/repodata/repomd.xml.key
enabled=1
[server-applications]
name=server-applications
type=rpm-md
baseurl=http://192.168.122.1/current_os/Module-Server-Applications/
gpgcheck=0
gpgkey=http://192.168.122.1/current_os/Module-Server-Applications/repodata/repomd.xml.key
enabled=1
[product-sles]
name=product-sles
type=rpm-md
baseurl=http://192.168.122.1/current_os/Product-SLES/
gpgcheck=0
gpgkey=http://192.168.122.1/current_os/Product-SLES/repodata/repomd.xml.key
enabled=1
current_ses.repo file:
# cat /srv/www/htdocs/current_ses.repo
[SES]
name=SES
type=rpm-md
baseurl=http://192.168.122.1/current_ses/
gpgcheck=0
gpgkey=http://192.168.122.1/current_ses/repodata/repomd.xml.key
enabled=1
systemctl enable docker libvirtd nfs-server apache2
systemctl start docker libvirtd nfs-server apache2Firewall should be disabled by default
systemctl disable SuSEfirewall2
systemctl stop SuSEfirewall2In some cases default Libvirt iptables rules are not desired. We are going to remove all REJECT rules.
iptables-save > /etc/iptables_orig.conf
cp /etc/iptables_orig.conf /etc/iptables.conf
sed -i '/REJECT/d' /etc/iptables.conf
iptables-resore /etc/iptables.conf
Vagrant is looking for storage-automation public and private keys in /root/.ssh/ path. If they are placed there we will be able to login without password to our deployed SES cluster. Thanks to vagrant-hostsupdater plugin we can use hostnames we defined in YAML file.
Pull from registry.suse.com a specific version of the container image called "registry":
docker pull registry.suse.com/sles12/registry:2.6.2
Create a directory for storing the imagesour container cache will know about:
mkdir -p /docker/perstorage/registry/data
Create config.yaml file one directory above:
# cat /docker/perstorage/registry/config.yml
version: 0.1
log:
fields:
service: registry
storage:
cache:
blobdescriptor: inmemory
filesystem:
rootdirectory: /var/lib/registry
maintenance:
readonly:
enabled: true
http:
addr: 0.0.0.0:5000
headers:
X-Content-Type-Options: [nosniff]
health:
storagedriver:
enabled: true
interval: 10s
threshold: 3
proxy:
remoteurl: https://registry.suse.de
Start container cache:
docker run -d --restart=always --name locreg -p 5000:5000 \
-v /docker/perstorage/registry/data:/var/lib/registry \
-v /docker/perstorage/registry/config.yml:/etc/docker/registry/config.yml \
-v /usr/share/pki/trust/anchors/SUSE_Trust_Root.crt.pem:/var/lib/ca-certificates/ca-bundle.pem \
registry.suse.com/sles12/registry:2.6.2 serve /etc/docker/registry/config.yml
Make sure that container is installed:
docker ps -a
Take a look into the container cache:
curl -s http://localhost:5000/v2/_catalog | jq
Pull container image from https://registry.suse.de
docker pull localhost:5000/suse/sle-15-sp2/update/products/ses7/milestones/containers/ses/7/ceph/ceph:latest
New image should be available:
docker image list
curl -s http://localhost:5000/v2/_catalog | jq
Vagrant is using your user home directory to store its data. In some cases it's not what we want. As it may eat all of your disk space.
Therefore we're going to move it under /qemu
mkdir /qemu/vagrant
rsync -avP $HOME/.vagrant.d /qemu/vagrant
Set up new Vagrant home in your ~/.bashrc
VAGRANT_HOME=/qemu/vagrant/.vagrant.d
Apply changes
source ~/.bashrc
Enable udp port in /etc/nfs.conf
udp=y
Restart nfs-server service
systemctl restart nfs-server
We need libvirt to be able to emulate aarch64 architecture.
Following steps needs to be done additionaly:
install these packages:
zypper in ovmf qemu-arm
in the folder libvirt_files_for_aarch64_emulation are files that we have to copy in to the following locations:
/usr/share/qemu/aavmf-aarch64-code.bin
/usr/share/qemu/aavmf-aarch64-vars.bin
/usr/share/qemu/firmware/60-aavmf-aarch64.jsonWe need to enable nvram support in QEMU
Edit file /etc/libvirt/qemu.conf and uncomment or add these lines
nvram = [
"/usr/share/qemu/ovmf-x86_64-ms-4m-code.bin:/usr/share/qemu/ovmf-x86_64-ms-4m-vars.bin",
"/usr/share/qemu/ovmf-x86_64-ms-code.bin:/usr/share/qemu/ovmf-x86_64-ms-vars.bin",
"/usr/share/qemu/aavmf-aarch64-code.bin:/usr/share/qemu/aavmf-aarch64-vars.bin"
]
Create directories for aarch64 repositories:
mkdir /srv/www/htdocs/current_aarch64_os
mkdir /srv/www/htdocs/current_aarch64_ses
Download and mount desired OS and SES versions:
wget -P /qemu/iso <OS ISO url>
mount <path to OS ISO> /srv/www/htdocs/current_aarch64_os
wget -P /qemu/iso <SES ISO url>
mount <path to SES ISO> /srv/www/htdocs/current_aarch64_ses
Create repo files for aarch64 under /srv/www/htdocs:
file current_aarch64_os.repo
[basesystem]
name=basesystem
type=rpm-md
baseurl=http://192.168.122.1/current_aarch64_os/Module-Basesystem/
gpgcheck=0
gpgkey=http://192.168.122.1/current_aarch64_os/Module-Basesystem/repodata/repomd.xml.key
enabled=1
[server-applications]
name=server-applications
type=rpm-md
baseurl=http://192.168.122.1/current_aarch64_os/Module-Server-Applications/
gpgcheck=0
gpgkey=http://192.168.122.1/current_aarch64_os/Module-Server-Applications/repodata/repomd.xml.key
enabled=1
[product-sles]
name=product-sles
type=rpm-md
baseurl=http://192.168.122.1/current_aarch64_os/Product-SLES/
gpgcheck=0
gpgkey=http://192.168.122.1/current_aarch64_os/Product-SLES/repodata/repomd.xml.key
enabled=1
file current_aarch64_ses.repo
[SES]
name=SES
type=rpm-md
baseurl=http://192.168.122.1/current_aarch64_ses/
gpgcheck=0
gpgkey=http://192.168.122.1/current_aarch64_ses/repodata/repomd.xml.key
enabled=1
Steps in this chapter are not needed if you are going to use build_validation.sh script. Script can do it for you automatically
Creating of the Vagrant box is automated with the build_validation.sh script.
In this example we are going to use SLE-15-SP2-Snapshot15 and SES7-M10.
We are not using prepared vagrant images but creating our own.
Download OS image:
mkdir /qemu/iso
wget http://download.suse.de/install/SLE-15-SP2-Full-Snapshot15/SLE-15-SP2-Full-x86_64-Snapshot15-Media1.iso -P /qemu/iso
Mount ISO under /srv/www/htdocs/current_os directory
mount /qemu/iso/SLE-15-SP2-Full-x86_64-Snapshot15-Media1.iso /srv/www/htdocs/current_os
Download SES image:
wget http://download.suse.de/install/SUSE-Enterprise-Storage-7-Milestone10/SUSE-Enterprise-Storage-7-DVD-x86_64-Milestone10-Media1.iso -P /qemu/iso
Mount ISO under /srv/www/htdocs/current_ses directory
mount /qemu/iso/SUSE-Enterprise-Storage-7-DVD-x86_64-Milestone10-Media1.iso /srv/www/htdocs/current_ses
Additionally for mixed architecture tests:
Mount ISO under /srv/www/htdocs/current_aarch64_os directory
mount /qemu/iso/SLE-15-SP2-Full-aarch64-Snapshot15-Media1.iso /srv/www/htdocs/current_aarch64_os
Mount ISO under /srv/www/htdocs/current_aarch64_ses directory
mount /qemu/iso/SUSE-Enterprise-Storage-7-DVD-aarch64-Milestone10-Media1.iso /srv/www/htdocs/current_ses
Insert autoyast file into Apache DocumetnRoot directory
cp autoyast/autoyast_intel.xml /srv/www/htdocs
Make sure OS image is mounted under /srv/www/htdocs/current_os
mount | grep current_os
Install QEMU VM with virt-install:
virt-install --name vgrbox --memory 2048 --vcpus 1 --hvm \
--disk bus=virtio,path=/qemu/pools/default/vgrbox.qcow2,cache=none,format=qcow2,size=10 \
--network bridge=virbr0,model=virtio --connect qemu:///system --os-type linux \
--os-variant sle15sp2 --virt-type kvm --noautoconsole --accelerate \
--location http://192.168.122.1/current_os \
--extra-args="console=tty0 console=ttyS0,115200n8 autoyast=http://192.168.122.1/autoyast_intel.xml"
Watch installation:
virsh console vgrbox
After VM reboots you need to start it manually:
virsh start vgrbox
Watch autoyast 2nd stage installation
virsh console vgrbox
After VM reboots second time DO NOT START IT AGAIN
Continue with Put image into the Vagrant boxes
Insert autoyast file into Apache DocumetnRoot directory
cp autoyast/autoyast_aarch64.xml /srv/www/htdocs
Make sure OS image is mounted under /srv/www/htdocs/current_os
mount | grep current_os
Install QEMU VM with virt-install:
virt-install --name vgrbox --memory 2048 --vcpus 1 --hvm \
--disk bus=virtio,path=/qemu/pools/default/vgrbox.qcow2,cache=none,format=qcow2,size=10 \
--network bridge=virbr0,model=virtio --connect qemu:///system --os-type linux \
--os-variant sle15sp2 --arch aarch64 --noautoconsole --accelerate \
--location http://192.168.122.1/current_os \
--extra-args="console=ttyAMA0,115200n8 autoyast=http://192.168.122.1/autoyast_aarch64.xml"
Watch installation:
virsh console vgrbox
After VM reboots you need to start it manually:
virsh start vgrbox
Watch autoyast 2nd stage installation
virsh console vgrbox
After VM reboots second time DO NOT START IT AGAIN
Continue with Put image into the Vagrant boxes
Make sure OS image is mounted under /srv/www/htdocs/current_os
mount | grep current_os
Install QEMU VM with virt-install:
virt-install --name vgrbox --memory 2048 --vcpus 1 --hvm \
--disk bus=virtio,path=/qemu/pools/default/vgrbox.qcow2,cache=none,format=qcow2,size=10 \
--network bridge=virbr0,model=virtio --connect qemu:///system --os-type linux \
--os-variant sle15sp2 --virt-type kvm --noautoconsole --accelerate \
--location http://192.168.122.1/current_os --extra-args="console=tty0 console=ttyS0,115200n8"
Watch installation:
virsh console vgrbox
After VM reboots you need to start it manually:
virsh start vgrbox
Login into the VM using console
virsh console vgrbox
Create bash file /tmp/vgrsetup.sh
set -ex
useradd -m vagrant
echo "vagrant ALL=(ALL) NOPASSWD:ALL" > /etc/sudoers.d/vagrant
chmod 600 /etc/sudoers.d/vagrant
mkdir -p /home/vagrant/.ssh
chmod 0700 /home/vagrant/.ssh
wget --no-check-certificate \
https://raw.github.com/mitchellh/vagrant/master/keys/vagrant.pub \
-O /home/vagrant/.ssh/authorized_keys
chmod 0600 /home/vagrant/.ssh/authorized_keys
chown -R vagrant /home/vagrant/.ssh
truncate -s 0 /etc/udev/rules.d/70-persistent-net.rules
rm -f /var/lib/wicked/* ~/.bash_history
poweroff
Run
while $(zypper lr 1 >/dev/null 2>&1); do zypper rr 1 >/dev/null 2>&1; done
Run our bash file
bash /tmp/vgrsetup.sh
After VM shutdown DO NOT START IT AGAIN
Continue with Put image into the Vagrant boxes
Make sure OS image is mounted under /srv/www/htdocs/current_os
mount | grep current_os
Install QEMU VM with virt-install:
virt-install --name vgrbox --memory 2048 --vcpus 1 --hvm \
--disk bus=virtio,path=/qemu/pools/default/vgrbox.qcow2,cache=none,format=qcow2,size=10 \
--network bridge=virbr0,model=virtio --connect qemu:///system --os-type linux \
--os-variant sle15sp2 --arch aarch64 --noautoconsole --accelerate \
--location http://192.168.122.1/current_os --extra-args="console=ttyAMA0,115200n8"
Watch installation:
virsh console vgrbox
After VM reboots you need to start it manually:
virsh start vgrbox
Login into the VM using console
virsh console vgrbox
Create bash file /tmp/vgrsetup.sh
set -ex
useradd -m vagrant
echo "vagrant ALL=(ALL) NOPASSWD:ALL" > /etc/sudoers.d/vagrant
chmod 600 /etc/sudoers.d/vagrant
mkdir -p /home/vagrant/.ssh
chmod 0700 /home/vagrant/.ssh
wget --no-check-certificate \
https://raw.github.com/mitchellh/vagrant/master/keys/vagrant.pub \
-O /home/vagrant/.ssh/authorized_keys
chmod 0600 /home/vagrant/.ssh/authorized_keys
chown -R vagrant /home/vagrant/.ssh
truncate -s 0 /etc/udev/rules.d/70-persistent-net.rules
rm -f /var/lib/wicked/* ~/.bash_history
poweroff
Create script /boot/efi/startup.nsh to start UEFI boot loader with path to your boot loader, mostly:
\EFI\sles\grubaa64.efi
Set up permissions:
chmod 644 /boot/efi/startup.nsh
Run
while $(zypper lr 1 >/dev/null 2>&1); do zypper rr 1 >/dev/null 2>&1; done
Run our bash file
bash /tmp/vgrsetup.sh
After VM shutdown DO NOT START IT AGAIN
Continue with Put image into the Vagrant boxes
Make sure OS image is mounted under /srv/www/htdocs/client
mount | grep client
Install QEMU VM with virt-install:
virt-install --name vgrbox --memory 2048 --vcpus 1 --hvm \
--disk bus=virtio,path=/qemu/pools/default/vgrbox.qcow2,cache=none,format=qcow2,size=10 \
--network bridge=virbr0,model=virtio --connect qemu:///system --os-type linux \
--os-variant sle15sp2 --virt-type kvm --noautoconsole --accelerate \
--location http://192.168.122.1/client --extra-args="console=tty0 console=ttyS0,115200n8"
Watch installation:
virsh console vgrbox
After VM reboots you need to start it manually:
virsh start vgrbox
Login into the VM using console
virsh console vgrbox
Some distributions like Fedora or RHEL or Ubuntu are not naming NIC with eth
We have to change it:
Edit /etc/default/grub file
GRUB_CMDLINE_LINUX_DEFAULT="maybe-ubiquity"
GRUB_CMDLINE_LINUX="net.ifnames=0 biosdevname=0"
Run
grub-mkconfig -o /boot/grub/grub.cfg
For Ubuntu 18.04 and higher change the file /etc/netplan/50-cloud-init.yaml to contain following:
network:
ethernets:
eth0:
addresses: []
dhcp4: true
optional: true
version: 2
Create bash file /tmp/vgrsetup.sh
set -ex
useradd -m vagrant
echo "vagrant ALL=(ALL) NOPASSWD:ALL" > /etc/sudoers.d/vagrant
chmod 600 /etc/sudoers.d/vagrant
mkdir -p /home/vagrant/.ssh
chmod 0700 /home/vagrant/.ssh
wget --no-check-certificate \
https://raw.github.com/mitchellh/vagrant/master/keys/vagrant.pub \
-O /home/vagrant/.ssh/authorized_keys
chmod 0600 /home/vagrant/.ssh/authorized_keys
chown -R vagrant /home/vagrant/.ssh
truncate -s 0 /etc/udev/rules.d/70-persistent-net.rules
rm -f /var/lib/wicked/* ~/.bash_history
poweroff
Run our bash file
bash /tmp/vgrsetup.sh
After VM shutdown DO NOT START IT AGAIN
Continue with Put image into the Vagrant boxes
Go into the Vagrant Home directory and create directory structure
mkdir -p /qemu/vagrant/.vagrant.d/boxes/sle15sp2snap15/0/libvirt
Create metadata file:
# cat /qemu/vagrant/.vagrant.d/boxes/sle15sp2snap15/0/libvirt/metadata.json
{
"provider" : "libvirt",
"format" : "qcow2",
"virtual_size" : 10
}
Create Vagrantfile:
# cat /qemu/vagrant/.vagrant.d/boxes/sle15sp2snap15/0/libvirt/Vagrantfile
Vagrant.configure("2") do |config|
config.vm.provider :libvirt do |libvirt|
libvirt.driver = "kvm"
libvirt.host = 'localhost'
libvirt.uri = 'qemu:///system'
end
config.vm.define "new" do |custombox|
custombox.vm.box = "custombox"
custombox.vm.provider :libvirt do |test|
test.memory = 1024
test.cpus = 1
end
end
end
Move our VM into this directory
mv /qemu/pools/default/vgrbox.qcow2 /qemu/vagrant/.vagrant.d/boxes/sle15sp2snap15/0/libvirt/box.img
Verify:
# vagrant box list
sle15sp2snap15
Undefine VM from libvirt DB
virsh undefine vgrbox
Create symbolic links (this is much faster and space saving if compared what Vagrant is doing)
ln -s /qemu/vagrant/.vagrant.d/boxes/sle15sp2snap15/0/libvirt/box.img \
/qemu/pools/default/sle15sp2snap15_vagrant_box_image_0.img
Restart libvirt daemon
systemctl restart libvirtd
There is a server ecp-registry.openstack.local (IP 10.86.1.251, login with storage-automation key) on OpenStack that runs Apache and Docker to provide repositories and Docker container cache with same setup as was described in
We need to patch cinder files to work with ECP.
Go to your vagrant home directory and run
cd .vagrant.d
find . -name cinder.rb
replace the file with cinder.rb
same for cinder_spec.rb
find . -name cinder_spec.rb
Download and source RC file from OpenStack.
Find existing JeOS image of required OS and create instance of it.
Download JeOS qcow2 file for example with name sle15sp2snap15-jeos.qcow2 to your local server.
Create image:
openstack image create --private --disk-format qcow2 --project-domain ses --file /qemu/pools/default/sle15sp2snap15-jeos.qcow2 sle15sp2snap15
We have to change some settings in image
Create instance
openstack server create --image sle15sp2snap15 --flavor m1.small --key-name storage-automation --availability-zone nova --network sesci sle15testimage
Get free floating IP and assign to instance
openstack floating ip list | grep -i none | tail -1
openstack server add floating ip sle15sp2testimage <IP>
Log in to the server as user sles and switch to root with sudo su -:
replace /etc/cloud/cloud.cfg file with this file
In user root home directory remove restrictions from beggining of line in /root/.ssh/authorized_keys file that it will looks like normal public key context.
execute commands:
systemctl restart cloud-init
netconf update -f
SUSEConnect -r <reg number>
zypper in -t pattern yast2_basis base enhanced_base
zypper in kernel-defaultCreate bash file /tmp/vgrsetup.sh
set -ex
useradd -m vagrant
echo "vagrant ALL=(ALL) NOPASSWD:ALL" > /etc/sudoers.d/vagrant
chmod 600 /etc/sudoers.d/vagrant
mkdir -p /home/vagrant/.ssh
chmod 0700 /home/vagrant/.ssh
wget --no-check-certificate \
https://raw.github.com/mitchellh/vagrant/master/keys/vagrant.pub \
-O /home/vagrant/.ssh/authorized_keys
chmod 0600 /home/vagrant/.ssh/authorized_keys
chown -R vagrant /home/vagrant/.ssh
truncate -s 0 /etc/udev/rules.d/70-persistent-net.rules
rm -f /var/lib/wicked/* ~/.bash_history
poweroff
Create new image from instance
nova image-create --show --poll sle15sp2snap15 <image name>
Try to create instance from new image and watch if it boots properly
Continue on server ecp-registry with steps described in Download and mount ISOs
Clone the vagrant_ses project.
Structure is according architecture we are going to deploy SES cluster and run tests on or there is openstack directory if we are going to deploy SES cluster on OpenStack.
./vagrant_ses/scripts directory contains bash scripts used for testing on SES6 and they are linked into appropriate folder.
For SES7 there is separate scripts directory under ./vagrant_ses/<architecture>/SES7/ or ./vagrant_ses/openstack/SES7/.
We are going to use SES7 and vagrant box sle15sp2snap15 as an example.
Edit YAML file with vagrant box for deployment:
ses_cl_box: 'sle15sp2snap15'If we are going to use build_validation.sh script then only these scripts needs to be enabled in YAML:
master_sh:
- hosts_file_correction.sh
- deploy_ses.shAll other scripts are handled by build_validation.sh script
In most cases nothing more needs to be edited.
Run test:
# ./build_validation.sh -h
usage: ./build_validation.sh --help
build_validation.sh [arguments]
arguments:
--vagrantfile VAGRANTFILE
--ses-only deploys only SES without running BV test scripts
--destroy destroys project (vagrant destroy -f)
--all-scripts runs all BV scripts under ./scripts directory
--only-script runs only specified script
--existing runs BV scripts on existing cluster
--only-salt-cluster deploys cluster with salt
--vagrant-box vagrant box name. Don't use with option --sle-slp-dir.
--sle-slp-dir directory of SLE Full SLP (example: SLE-15-SP2-Full-Snapshot16)
--ses-slp-dir directory of SES SLP (example: SUSE-Enterprise-Storage-7-Milestone11)
--destroy-before-deploy destroys existing cluster before deployment (useful for Jenkins)
Example:
./build_validation.sh --vagrantfile SES7_Build_validation --destroy-before-deploy \
--all-scripts --sle-slp-dir SLE-15-SP2-Full-Snapshot16 \
--ses-slp-dir SUSE-Enterprise-Storage-7-Milestone11
What is script doing:
-
deploys SES cluster with Vagrant using repositories we did set up in Download and mount ISOs or if we selected SCC in YAML then it using these repositories for deploying SES and for OS. We also can use these options together if we need to test some additional packages that are not available in SCC.
-
creates snapshot deployment of fresh deployed cluster
-
if some script fails then
-
it collects supportconfig logs from whole cluster and copies them to ./logs/ directory
-
creates snapshot of cluster with name of the script that failed
-
reverts to fresh deployed cluster snapshot
-
-
reverts to fresh deployed cluster snapshot after each script is finished
We can manage cluster also with virsh tool
# virsh list
Id Name State
--------------------------------
33 SES7_master running
34 SES7_osd-node1 running
35 SES7_osd-node2 running
36 SES7_osd-node3 running
37 SES7_osd-node4 running
38 SES7_osd-node5 running
39 SES7_monitor1 running
40 SES7_monitor2 running
41 SES7_monitor3 running
# virsh snapshot-list SES7_master
Name Creation Time State
---------------------------------------------------
deployment 2020-05-14 13:24:18 +0200 shutoff
Our project directory (where Vagrant files and scripts folder are located) is mounted over NFS to each server under /vagrant directory.
Edit YAML file as described in Running tests with build validation script
and additionally enable scripts you want Vagrant to run on SES cluster. They will run in order as they are in YAML file.
Disadvantage is that Build Validation test will fail if some of the scripts will fail and cluster may stay in broken state. You will need to destroy it and run again.
Run tests
VAGRANT_VAGRANTFILE=SES7_Build_Validation vagrant up
Destroy cluster
VAGRANT_VAGRANTFILE=SES7_Build_Validation vagrant destroy -f
Vagrant can't destroy libvirt domain as it is using nvram.
Use destroy_project.sh
./destroy_project.sh <Vagrantfile>
Go into the folder ./vagrant_ses/openstack/SES7
Download RC file from OpenStack and source it.
source ses-openrc.sh
Edit YAML file as described in Running tests just with Vagrant where ses_cl_box is the name of the image.
Run
python prepare_openstack_for_vagrant.py -y SES7_Build_validation_OpenStack.yaml -f SES7_Build_validation_OpenStack
it will delete servers definded in YAML file from OpenStack if they exists and delete "vgr_volume*" and creates new ones
run
VAGRANT_VAGRANTFILE=SES7_Build_validation_OpenStack vagarant up --provider=openstack