# Ubuntu/Debian based Openstack enviroments

## **Generic Pre-requisites**

1. Please ensure the following points before starting the upgrade process:
   1. *No snapshot OR restore to be running.*
   2. *Global job scheduler should be disabled.*
2. The mentioned gemfury repository should be accessible from TVault VM.

## **Rolling upgrade on Openstack Ansible**

### Repo Updates

Add gemfury repo on each dmapi, horizon containers & compute node(s) to get updated packages.

Create a file /etc/apt/sources.list.d/fury.list and add below line in it.

```
deb [trusted=yes] https://apt.fury.io/triliodata-4-0/ /
```

Use the below commands to get a list of updated packages available on the configured repositories.

```
apt-get update
apt list --upgradable 
```

### Upgrade tvault-datamover-api package

* Login to dmapi container from the controller node using the below command.

```
lxc-attach -n controller_dmapi
```

* Take a backup of the following file on each dmapi container(s).

```
tar -czvf dmapi_config.tar.gz /etc/dmapi
```

* Add the gemfury repo and upgrade the dmapi package using the below command. Select the appropriate package depending on the python version used.

```
apt install python3-dmapi --upgrade
```

* Restore the backed-up config files

```
tar -xzvf dmapi_config.tar.gz -C /
```

* Now restart and check the service tvault-datamover-api on controller.

```
systemctl restart tvault-datamover-api
systemctl status tvault-datamover-api
```

### Upgrade Horizon plugin

* For the Horizon plugin upgrade, we need to upgrade below two packages.

python3-tvault-horizon-plugin\
python3-workloadmgrclient

* Login to the horizon container from the controller node using the below command.

  ```
  lxc-attach -n controller_horizon_container-ead7cc60
  ```
* Add the gemfury repo and upgrade the tvault-horizon-plugin & workloadmgrclient packages using the below command. Select the appropriate package depending on the python version used.

```
apt install python3-tvault-horizon-plugin --upgrade
apt install python3-workloadmgrclient --upgrade
```

* Restart the apache2 service and verify the workloadmgrclient version using the below commands.

```
systemctl restart apache2
workloadmgr --version
```

### Upgrade the tvault-contego package

#### **NFS Storage Backend**

* Take a backup of following file on each compute node(s) for nfs storage backend.

```
tar -czvf  contego_config.tar.gz /etc/tvault-contego/ 
```

* Upgrade the tvault-contego package using below command.Select the appropriate package depending on python version used.

```
apt install python3-tvault-contego --upgrade
```

* Restore the backed-up config files

```
 tar -xzvf  contego_config.tar.gz -C /
```

* Now restart and check the service tvault-contego on compute node(s).

```
systemctl restart tvault-contego
systemctl status tvault-contego
```

Please make sure all configuration files are unchanged.

#### **S3 Storage Backend**

* Take a backup of following file on each compute node(s) for s3 storage backend.

```
tar -czvf  contego_config.tar.gz /etc/tvault-contego/ /etc/tvault-object-store/
```

* Upgrade the tvault-contego package using below command. Select the appropriate package depending on the python version used.

```
apt install python3-tvault-contego python3-s3-fuse-plugin --upgrade
```

Note\*:- If you get prompt while installing, choose the default selection.

* Restore the backed-up config files

```
 tar -xzvf  contego_config.tar.gz -C /
```

* Now restart and check the service tvault-contego on compute node(s).

```
systemctl restart tvault-object-store
systemctl restart tvault-contego 
systemctl status tvault-object-store
systemctl status tvault-contego
```

Please make sure all configuration files are unchanged.

## **Rolling upgrade on Kolla Openstack**

### Upgrade tvault-datamover-api package (*trilio\_datamover\_api Container*)

Container trilio\_datamover\_api needs to be redeployed. Follow below steps on **all controller nodes** :

* Backup Existing conf files/folders

```
Copy folders /etc/kolla/nova-api & /etc/kolla/trilio-datamover-api/ to some local location.
Eg. cp -r /etc/kolla/nova-api $HOME/4.0-BK/
```

* Stop and Remove existing container trilio\_datamover\_api

```
docker stop trilio_datamover_api
docker rm trilio_datamover_api
```

* Pull trilio\_datamover\_api container image:

```
## Login to docker using credentials: triliodocker/triliopassword
docker login docker.io

## Pull Trilio Datamover Api container using following command. You need to edit OS_PLATFORM, TVAULT_VERSION,
 OPENSTACK_RELEASE_NAME.

<OS_PLATFORM> could be ubuntu/centos.
<OPENSTACK_RELEASE_NAME> could would be train
<TVAULT_VERSION>    : For exact GA build version check section 1. i) Plan for deployment

docker pull docker.io/trilio/<OS_PLATFORM>-source-trilio-datamover-api:<TVAULT_VERSION>-<OPENSTACK_RELEASE_NAME>

Example command for train openstack on ubuntu platform with triliovault 4.0 release:
docker pull docker.io/trilio/ubuntu-source-trilio-datamover-api:4.0.106-train
```

* Run datamover api container.

```
You need to edit OS_PLATFORM, TVAULT_VERSION, OPENSTACK_RELEASE_NAME. 
<OS_PLATFORM> could be ubuntu/centos.
<OPENSTACK_RELEASE_NAME> would be train
<TVAULT_VERSION> : For exact TVAULT build version check section 1. i) Plan for deployment

docker run --network host --name trilio_datamover_api -d --restart always -v /etc/kolla/trilio-datamover-api/nova.conf:/etc/nova/nova.conf \
-v /etc/kolla/trilio-datamover-api/dmapi.conf:/etc/dmapi/dmapi.conf \
-v /var/log/kolla/:/var/log/kolla/ \
trilio/<OS_PLATFORM>-source-trilio-datamover-api:<TVAULT_VERSION>-<OPENSTACK_RELEASE_NAME> /var/lib/kolla/venv/bin/python /usr/bin/dmapi-api 
```

* Verify deployment

```
Make sure 'trilio_datamover_api' named container running well and not in 'restarting' state on all controller nodes. 
'docker ps' command can be used for the same.

# docker ps | grep trilio_datamover_api
e603ea78094f  trilio/ubuntu-source-trilio-datamover-api:4.0.106-train   "dumb-init --single-…"   5 minutes ago Up 5 minutes
trilio_datamover_api
```

### **Upgrade Horizon Plugin Package**

* Login to horizon container (***on controller node***)

```
docker exec -it horizon /bin/bash
```

* Add gemfury repo on horizon to get updated packages.

  Create a file /etc/apt/sources.list.d/fury.list and add below line in it.

  ```
  deb [trusted=yes] https://apt.fury.io/triliodata-4-0/ /
  ```

  Use below commands to get list of updated packages available on repo's.

  ```
  apt-get update
  apt list --upgradable
  ```
* For Horizon plugin upgrade, following packages need to be upgraded.

*python3-tvault-horizon-plugin*\
\&#xNAN;*python3-workloadmgrclient*

* Run following commands to upgrade the tvault-horizon-plugin & workloadmgrclient packages. *Select the appropriate package depending on python version used.*

```
dpkg -l | grep -i tvault
  ii  python3-tvault-horizon-plugin  4.0.92    all   tVault plugin for  OpenStack Horizon Dashboard
  ii  python3-workloadmgrclient      4.0.92   all  tVault Workload Manager Client

apt install python3-tvault-horizon-plugin --upgrade
apt install python3-workloadmgrclient --upgrade

dpkg -l | grep -i tvault
   ii  python3-tvault-horizon-plugin  4.0.115                             all          tVault plugin for  OpenStack Horizon Dashboard
   ii  python3-workloadmgrclient      4.0.115                             all          tVault Workload Manager Client
```

* Restart docker container (*from controller node*) and verify the workloadmgrclient version (*inside horizon container*).

```
docker restart horizon
docker exec -it horizon /bin/bash
workloadmgr --version
```

### Upgrade Tvault-Contego (*trilio\_datamover Container*)

Container *trilio\_datamover* needs to be redeployed. Follow below steps on **all Compute nodes** :

* Backup Existing conf files/folders

```
Copy folder /etc/kolla/trilio-datamover to some local location.
Eg. cp -r /etc/kolla/trilio-datamover $HOME/4.0-BK/
```

* Stop & remove existing trilio\_datamover container

```
docker stop trilio_datamover
docker rm trilio_datamover
```

* Pull Trilio Datamover container image using the following command:

```
## Login to docker using credentials: triliodocker/triliopassword
docker login docker.io

## Pull Trilio Datamover Api container using following command. You need to edit OS_PLATFORM, TVAULT_VERSION,
OPENSTACK_RELEASE_NAME.

<OS_PLATFORM> could be ubuntu/centos.
<OPENSTACK_RELEASE_NAME> would be train
<TVAULT_VERSION> : For exact TVAULT build version check section 1. i) Plan for deployment

docker pull docker.io/trilio/<OS_PLATFORM>-source-trilio-datamover:<TVAULT_VERSION>-<OPENSTACK_RELEASE_NAME>

Example command for train openstack on ubuntu platform with Trilio 4.0 release:
docker pull docker.io/trilio/ubuntu-source-trilio-datamover:4.0.106-train
```

* Run datamover container.

```
-- In following docker run command you need to edit OS_PLATFORM, TVAULT_VERSION,
OPENSTACK_RELEASE_NAME.

<OS_PLATFORM> could be ubuntu/centos.
<OPENSTACK_RELEASE_NAME> would be train
<TVAULT_VERSION> : For exact TVAULT build version check section 1. i) Plan for deployment

Note: If your cloud does not use 'ceph' storage for nova/cinder, remove '/etc/ceph' volume mount option 
from below commands.

##If backup storage type is 'NFS'
docker run --privileged --network host --name trilio_datamover -d \
--restart always \
-v /etc/kolla/trilio-datamover/tvault-contego.conf:/etc/tvault-contego/tvault-contego.conf \
-v /etc/kolla/trilio-datamover/nova.conf:/etc/nova/nova.conf \
-v /dev:/dev:rw \
-v /etc/ceph:/etc/ceph:ro \
-v nova_compute:/var/lib/nova/:rw \
-v /var/log/kolla/:/var/log/kolla/ \
-v iscsi_info:/etc/iscsi:rw -v /var/run/libvirt \
trilio/<OS_PLATFORM>-source-trilio-datamover:<TVAULT_VERSION>-<OPENSTACK_RELEASE_NAME> /opt/tvault/start_datamover_nfs
```

* Verify Deployment of trilio\_datamover

```
Make sure 'trilio_datamover' named container running well and not in 'restarting' state on all compute nodes.
   'docker ps' command can be used for the same.
# docker ps | grep trilio_datamover
3e62813e5ade        trilio/ubuntu-source-trilio-datamover:4.0.115-train   "dumb-init --single-…"   7 hours ago         Up 7 hours                              trilio_datamover
```