ステージングインスタンスのセットアップ

About staging instances

GitHub recommends that you set up a separate environment to test backups, updates, or changes to the configuration for your GitHub Enterprise Server instance. This environment, which you should isolate from your production systems, is called a staging environment.

For example, to protect against loss of data, you can regularly validate the backup of your production instance. You can regularly restore the backup of your production data to a separate GitHub Enterprise Server instance in a staging environment. On this staging instance, you could also test the upgrade to the latest feature release of GitHub Enterprise Server.

Considerations for a staging environment

To thoroughly test GitHub Enterprise Server and recreate an environment that's as similar to your production environment as possible, consider the external systems that interact with your instance. For example, you may want to test the following in your staging environment.

• Authentication, especially if you use an external authentication provider like SAML
• Integration with an external ticketing system
• Integration with a continuous integration server
• External scripts or software that use the GitHub Enterprise Server APIs
• External SMTP server for email notifications

Setting up a staging instance

You can set up a staging instance from scratch and configure the instance however you like. For more information, see "Setting up a GitHub Enterprise Server instance" and "Configuring GitHub Enterprise."

Alternatively, you can create a staging instance that reflects your production configuration by restoring a backup of your production instance to the staging instance.

1. Back up your production instance.
2. Set up a staging instance.
3. Configure GitHub Actions.
4. Configure GitHub Packages.
5. Restore your production backup.
6. Review the instance's configuration.
7. Apply the instance's configuration.

1. Back up your production instance

If you want to test changes on an instance that contains the same data and configuration as your production instance, back up the data and configuration from the production instance using GitHub Enterprise Server Backup Utilities. For more information, see "Configuring backups on your instance."

2. Set up a staging instance

Set up a new instance to act as your staging environment. You can use the same guides for provisioning and installing your staging instance as you did for your production instance. For more information, see "Setting up a GitHub Enterprise Server instance."

If you plan to restore a backup of your production instance, continue to the next step. Alternatively, you can configure the instance manually and skip the following steps.

3. Configure GitHub Actions

Optionally, if you use GitHub Actions on your production instance, configure the feature on the staging instance before restoring your production backup. If you don't use GitHub Actions, skip to "1. Configure GitHub Packages."

1. SSH into the staging instance. For more information, see "Accessing the administrative shell (SSH)."

   Shell

   ssh -p 122 admin@HOSTNAME
   

2. To configure the staging instance to use an external storage provider for GitHub Actions, enter one of the following commands.

   • Azure Blob Storage:

     Shell

     ghe-config secrets.actions.storage.blob-provider "azure"
     

   • Amazon S3:

     Shell

     ghe-config secrets.actions.storage.blob-provider "s3"
     

   • Google Cloud Storage:

     Shell

     ghe-config secrets.actions.storage.blob-provider "gcs"
     

3. Configure the external storage connection by entering the following commands, replacing the placeholder values with actual values for your connection.

   • Azure Blob Storage:

     Shell

     ghe-config secrets.actions.storage.azure.connection-string "CONNECTION STRING"
     

   • Amazon S3:

     Shell

     ghe-config secrets.actions.storage.s3.bucket-name "S3 BUCKET NAME"
     ghe-config secrets.actions.storage.s3.service-url "S3 SERVICE URL"
     ghe-config secrets.actions.storage.s3.access-key-id "S3 ACCESS KEY ID"
     ghe-config secrets.actions.storage.s3.access-secret "S3 ACCESS SECRET"
     

     Optionally, to force path-style addressing for S3, also enter the following command.

     Shell

     ghe-config secrets.actions.storage.s3.force-path-style true
     

   • Google Cloud Storage:

     Shell

     ghe-config secrets.actions.storage.gcs.service-url "SERVICE URL"
     ghe-config secrets.actions.storage.gcs.bucket-name "BUCKET NAME"
     ghe-config secrets.actions.storage.gcs.access-key-id "HMAC ACCESS ID"
     ghe-config secrets.actions.storage.gcs.access-secret "HMAC SECRET"
     

4. To prepare to enable GitHub Actions on the staging instance, enter the following command.

   Shell

   ghe-config app.actions.enabled true
   

4. Configure GitHub Packages

Optionally, if you use GitHub Packages on your production instance, configure the feature on the staging instance before restoring your production backup. If you don't use GitHub Packages, skip to "1. Restore your production backup."

1. Review the backup you will restore to the staging instance.

   • If you took the backup with GitHub Enterprise Server Backup Utilities 3.5 or later, the backup includes the configuration for GitHub Packages. Continue to the next step.
   • If you took the backup with GitHub Enterprise Server Backup Utilities 3.4 or earlier, configure GitHub Packages on the staging instance. For more information, see "Getting started with GitHub Packages for your enterprise."

2. SSH into the staging instance. For more information, see "Accessing the administrative shell (SSH)."

   Shell

   ssh -p 122 admin@HOSTNAME
   

3. Configure the external storage connection by entering the following commands, replacing the placeholder values with actual values for your connection.

   • Azure Blob Storage:

     Shell

     ghe-config secrets.packages.blob-storage-type "azure"
     ghe-config secrets.packages.azure-container-name "AZURE CONTAINER NAME"
     ghe-config secrets.packages.azure-connection-string "CONNECTION STRING"
     

   • Amazon S3:

     Shell

     ghe-config secrets.packages.blob-storage-type "s3"
     ghe-config secrets.packages.service-url "S3 SERVICE URL"
     ghe-config secrets.packages.s3-bucket "S3 BUCKET NAME"
     ghe-config secrets.packages.aws-access-key "S3 ACCESS KEY ID"
     ghe-config secrets.packages.aws-secret-key "S3 ACCESS SECRET"
     

4. To prepare to enable GitHub Packages on the staging instance, enter the following command.

   Shell

   ghe-config app.packages.enabled true
   

5. Restore your production backup

Use the ghe-restore command to restore the rest of the data from the backup. For more information, see "Configuring backups on your instance."

If the staging instance is already configured and you want to overwrite settings, certificate, and license data, add the -c option to the command. For more information about the option, see Using the backup and restore commands in the GitHub Enterprise Server Backup Utilities documentation.

6. Review the instance's configuration

To access the staging instance using the same hostname, update your local hosts file to resolve the staging instance's hostname by IP address by editing the /etc/hosts file in macOS or Linux, or the C:\Windows\system32\drivers\etc file in Windows.

Then, review the staging instance's configuration in the Management Console. For more information, see "Administering your instance from the web UI."

7. Apply the instance's configuration

To apply the configuration from the Management Console, click Save settings.

Bringing a staging instance back online

You may want to power off a staging instance to save costs and power it back on when needed.

An instance can stay offline for 60 days.

If you bring the instance back online within the allowed offline time period, GitHub Enterprise Server instantiates successfully. If the instance stays offline for longer than the allowed period, GitHub Enterprise Server fails to instantiate successfully, and an error message with the text server has been offline for more than the configured server_rejoin_age_max may appear in the system logs. See "About system logs."

If the instance is stuck in the error state, you can run these commands to recover.

sudo mv /data/user/consul/server_metadata.json /data/user/consul/server_metadata.json.bak
ghe-config-apply

Further reading

• "About upgrades to new releases"