# Completing Major Upgrades in TVE/Partner Environments

- Published on Sep 18, 2024
- 5 minute(s) read

Prior to v30, Hyperscience was shipped with a PostgreSQL 9.5.14 database. However, Hyperscience versions v30 and later come with a PostgreSQL 12.5 database, which may cause issues when upgrading from v28 to v30. Because we use the same PostgreSQL versions in TVE installations, the same issues may occur when upgrading TVE instances.

This article describes how to perform an upgrade of TVE/partner environments running Hyperscience v28 and earlier, and it provides the tools necessary to complete the process.

When upgrading, you can keep the current local database of your environment, or you can backup its content and restore it to the database included in the latest application version.

Therefore, this article covers two scenarios:

- A major upgrade without replacing the current local database with a new one
- A major upgrade that includes a database upgrade and the backing up and restoring of its contents

## Upgrading Hyperscience without replacing the current local database with a new one

> We strongly recommend creating a backup of your database before upgrading Hyperscience.

In some cases, you might need to keep the current database and not upgrade it. Other cases might require you to offload the data to an external database.

To upgrade a TVE/partner environment to a newer version (e.g., from v28 to v30) while keeping the database container intact, do the following:

### 1. Download the new bundle

Currently, this bundle is the v30 bundle.

### 2. Extract the bundle

```bash
tar xzvf <PATH_TO_NEW_BUNDLE>.tgz
```

### 3. Go to the directory where the v30 bundle was extracted

```powershell
cd <PATH_TO_NEW_BUNDLE>
```

### 4. Copy the “.env” file from the previous bundle to the directory of the new bundle

Provided that your current directory is , run the following to copy the ".env" file there:

```powershell
cp <PATH_TO_PREVIOUS_BUNDLE>/.env .
```

### 5. Start the application the standard way

```bash
sudo bash run.sh init
sudo bash run.sh
```

The PostgreSQL container won’t be changed, and the version used will be from the previous Hyperscience bundle (i.e., 9.5.14).

## Upgrading Hyperscience and upgrading the database to а new version

These instructions describe how to upgrade the local database, which we recommend doing before any major upgrade.

Hyperscience v30 comes with a new version of the local PostgreSQL image. The steps below describe how to migrate the content of the current local database to the new database.

The following instructions rely on a set of additional scripts, attached to this article:

- common.sh
- postgres_backup.sh
- postgres_restore.sh

Specific instructions about the scripts can be found by opening them and reading their headers.

### 1. Download the new bundle

### 2. Extract the bundle in a directory

```powershell
tar xzvf <PATH_TO_NEW_BUNDLE>.tgz
```

### 3. Copy the “.env” file from the previous bundle directory to the new bundle directory

```powershell
cp /.env <PATH_TO_NEW_BUNDLE>/.env
```

### 4. Download the three scripts from the bottom of this article, and put them in the environment's new bundle directory

The scripts need to be present in the newly extracted bundle directory because _postgres_restore.sh_ has to be alongside the newest _postgres.tar_ coming with the new bundle.

### 5. Make the scripts executable

```powershell
chmod +x <PATH_TO_NEW_BUNDLE>/common.sh
chmod +x <PATH_TO_NEW_BUNDLE>/postgres_backup.sh
chmod +x <PATH_TO_NEW_BUNDLE>/postgres_restore.sh
```

Note that you can run Docker commands on Podman with the _podman-docker_ package. To learn how to install the _podman-docker_ package, see [step 2 of Configuring Podman in RHEL 8](https://help.hyperscience.ai/deployment/docs/configuring-podman-in-rhel-8#configuring-podman).

### 6. Run the _postgres_backup.sh_ script by passing it the “.env” file that was copied to the new bundle directory

```powershell
cd <PATH_TO_NEW_BUNDLE>
./postgres_backup.sh .env
```

The last command will create a backup of your current local database in the form of a SQL file. This file can also be used to restore the database contents to a third-party database. In the example below, the backup file is named _dbbackupfile-2021-06-25-07:32:09.sql_.

Remember the name of this backup file. This information will be part of the script's _stdout_.

### 7. Run the _postgres_restore.sh_ script by passing it the SQL backup file that was created in the previous step

Running this script will convert the data from a format that is readable by the current PostgreSQL database (presumably v9.5.14) to a format readable in v12.5.

```powershell
cd
./postgres_restore.sh .env /mnt/hs/db_backup/dbbackupfile-2021-06-25-07:32:09.sql
```

### 8. Start the new database from the latest bundle

```bash
cd
sudo bash run.sh db
```

### 9. Start the application the standard way

```bash
sudo bash run.sh init
sudo bash run.sh
```