blob: f9368d2020798afc6ebd57d794af00fbd1d33096 [file] [log] [blame]
How to deploy TryMOS on AWS
===========================
1. Dowload tryMOS image from Mirantis repo
wget https://artifactory.mcp.mirantis.net/artifactory/binary-dev-kaas-local/trymos/bin/trymos-bionic-amd64-master-20210316183204.qcow2
2. Convert QCOW2 image to RAW format
qemu-img convert -f qcow2 -O raw ./trymos-bionic-amd64-master-20210316183204.qcow2 ./trymos-bionic-amd64-master-20210316183204.raw
3. Upload RAW image to S3 storage, where `trymos-raw` name of a bucket
aws s3 cp ./trymos-bionic-amd64-master-20210316183204.raw s3://trymos-raw
4. Create a snapshot from the image which had been uploaded
aws ec2 import-snapshot --disk-container file://containers.json
containers.json
{
"Description": "TryMOS RAW",
"Format": "RAW",
"UserBucket": {
"S3Bucket": "trymos-raw",
"S3Key": "trymos-bionic-amd64-master-20210316183204.raw"
}
}
5. Create an image from the snapshot in EC2 Service -> Elastic Block Store -> Snapshots -> Actions -> Create image
with root storage 30Gb and additional volume 20Gb (volume type EBS (gp3))
6. Launch instance from the image EC2 Service -> Images -> AMIs -> Launch with flavor minimal 16 CPUs and
30Gb RAM (c4.4xlarge).
7. Connect to intance via external IP with keyfile which was defined during instance was brought up as root
ssh 13.59.177.99 -i ./trymos-pair.pem -l root
8. Run following commands to prepare env for deployment
sed -i 's/vdb/xvdb/g' /usr/share/metadata/ceph.yaml
and export control network CIDR
root@ip-172-31-11-89:/srv/bin# ip route
default via 172.31.0.1 dev eth0 proto dhcp src 172.31.11.89 metric 100
172.31.0.0/20 dev eth0 proto kernel scope link src 172.31.11.89
172.31.0.1 dev eth0 proto dhcp scope link src 172.31.11.89 metric 100
root@ip-172-31-11-89:/srv/bin# export CONTROL_NETWORK_CIDR='172.31.0.0/20'
9. Run installation script
/srv/bin/bootstrap_trymos_aws.sh