Features

With the new features added in GitHub Enterprise 2.9.0, you can:

• Configure the SAML username attribute name to match your organizational standard or what your identity provider uses.
• Disable password-based authentication for Git operations when using LDAP authentication.
• Use load balancers for a single appliance of GitHub Enterprise and in front of a High Availability environment.
• Use TLS encrypted communication for secure log forwarding.
• Deploy GitHub Enterprise on Google Compute Engine.
• Require users to download recovery codes before they can complete the 2FA setup process.
• Manage one or more users at a time with bulk management.
• Remove one or more outside collaborator's access to a single repository or to all organization owned repositories at one time.
• Locate users by their GPG key from /stafftools.
• Purge LFS objects in a repository from /stafftools.
• Search commits by fields such as message, author, and date.
  • Note: Due to indexing commits, this feature may require adding more space to the data partition (depending on the number of commits and the length of each message).
• View diffs beyond the previous 300 file maximum.
• Resolve merge conflicts from your pull requests.
• Request a pull request review from a specific person.
• Dismiss a review from a pull request.
• Preserve your forks even if you are removed from a team using LDAP sync.
• Identify whether a user is a contributor, owner, or member from badges in the issue or pull request comment.
• Allow organizations to pin repositories.
• Create an organization-wide project to manage issues and pull requests from any repository that belongs to an organization.
• Resume downloads of Git LFS objects, releases, and uploads with supported HTTP(S) clients.

Security Fixes

• Packages have been updated to the latest security versions.

Bug Fixes

• Directory names containing spaces failed to be created from the web interface.
• Directional hotkeys weren't functional on the commit activity graph.
• Searching repositories from the organization page was inconsistent with the results from the global search.
• Commits were incorrectly referenced to the parent repository in the timeline after restoring a fork.
• Validly signed commits were displayed as invalid if the client added metadata to the gpg signature.
• Attempting to convert a user to an organization failed with an internal server error.
• Git LFS objects could take up to an hour to replicate in a High Availability configuration.
• In a clustering environment, reindexing failed when a pull request routes to an offline repository.
• svn checkout could incorrectly fail with a 429 HTTP error code.
• Copy to clipboard buttons failed for Internet Explorer 11 users.
• Pre-receive hooks failed to output UTF-8 characters.
• Migrations failed to preserve a label with a / character.
• A previously configured replica appliance excessively logged errors during High Availability initialization.
• Pre-receive hooks could fail with Device or resource busy.
• LDAP synchronization incorrectly removed users after a server-side LDAP timeout.
• An Encoding::Compatibility error occurred when viewing a webhook from /stafftools.
• The Management Console Add new SSH key field incorrectly allowed an SSH fingerprint instead of the contents of the key.
• A former primary appliance failed to create or update pre-receive hook environments.
• In a clustering environment, services were incorrectly started after reboot.
• Pre-receive hooks checking internal or temporary Git references failed.
• In a clustering environment, releases, uploads, avatars, and LFS files could fail to be accessible after a storage-server is removed.
• In a clustering environment, storage-server repair jobs took a long time when a new storage-server is added.
• In a clustering environment, the enterprise-manage and resolvconf service were incorrectly stopped after ghe-cluster-config-apply.
• An updated SAML Verification certificate did not take effect until the github-unicorn service was restarted.

Changes

• The Reactivate suspended users configuration has changed to reflect the current configured state.
• The <Destination> element is no longer optional in the SAML response.
• The default Amazon Web Services EC2 root partition has increased to 80 GB.
• GitHub Flavored Markdown, which is used to render issue and pull request comments, is now compliant with CommonMark.
  • Note: Rendering for repository files has not changed.
• New webhook events have been added.
• New API resources have been added.
• High Availability Git replication has been updated to use Spokes.
• A ten second timeout is enforced for all LDAP authentication requests. In the event of a timeout, the user is notified, and the timeout is recorded to the log files and reflected on the LDAP Authentication Management Console monitoring graph.
• Outdated diffs and review comments are now hidden by default when viewing a Pull Request. (updated 2017-05-17)

Backups and Disaster Recovery

GitHub Enterprise 2.9 requires at least GitHub Enterprise Backup Utilities 2.9.0 for Backups and Disaster Recovery.

Deprecation of GitHub Enterprise 2.4

GitHub Enterprise 2.4 is now deprecated as of February 9, 2017. That means that no patch releases will be made, even for critical security issues, after this release. For better performance, improved security, and new features, upgrade to the newest version of GitHub Enterprise as soon as possible.

Upcoming deprecation of GitHub Enterprise 2.5

GitHub Enterprise 2.5 will be deprecated as of March 14, 2017. That means that no patch releases will be made, even for critical security issues, after this date. For better performance, improved security, and new features, upgrade to the newest version of GitHub Enterprise as soon as possible.

Known Issues

• We incorrectly redirect to the dashboard if you accessed GitHub Enterprise using an alias while in private mode. This might happen if you set a fully qualified domain name but the subdomain resolves correctly.
• Images uploaded to issues save with an absolute URL, so they can be broken if the hostname changes.
• On a freshly set up GitHub Enterprise without any users, an attacker could create the first admin user.
• Custom firewall rules aren't maintained during an upgrade.
• Enqueued background jobs are sometimes not purged when a repository is deleted.
• svn checkout may timeout while the repository data cache is being built. In most cases, subsequent svn checkout attempts will succeed.
• Git LFS tracked files uploaded through the web interface are incorrectly added directly to the repository.
• GitHub Enterprise clustering can not be configured without https.
• Graphs in the Management Console monitoring page are incorrectly sorted.
• Site administrators may experience 500 Internal Server Error if the license is approaching expiration or is close to the seat limit. (updated 2017-03-08)
• The initial import of the VMware OVA image may fail when deployed via vCenter Server 6.0 or 6.5. The import will succeed when performed directly on an ESXi host.
• On Google Compute Engine, it's possible to use an ephemeral scratch disk as repository storage (updated 2017-03-10)
• A search index that's not marked as the primary index, for example when a new index is being built after an upgrade, can be incorrectly deleted. (updated 2017-03-10)
• It's possible to queue more jobs to repair a search index through the site admin than can be processed in a reasonable time, causing low priority jobs to become backlogged. (updated 2017-03-10)
• Deleting a search index doesn't delete all associated metadata, which is then incorrectly reused if a new search index is created. This can cause search index repair jobs to be reported as finished in the site admin when they're not. (updated 2017-03-10)
• An unused locations search index is incorrectly listed in the site admin indexing page. (updated 2017-03-10)
• It's not possible to enable or disable maintenance mode through the Management Console. Maintenance mode can still be enabled and disabled using the ghe-maintenance command line utility. (updated 2017-03-10)
• Accessing a GitHub Pages site could cause 500 Internal Server Error. (updated 2017-03-10)
• A configuration run can incorrectly revert an SSL certificate to an automatically generated self-signed certificate. (updated 2017-03-10)
• Starting high availability replication can fail if the appliance was previously configured as a replica. (updated 2017-03-10)
• The /status endpoint on a high availability replica incorrectly returns 200 OK instead of 503 Service Unavailable.
• Site administrators can experience a 500 Internal Server Error after viewing the history for a file path containing Japanese characters. (updated 2017-03-30)
• An issue or pull request comment containing the string "User-Agent: GitHub-Hookshot" incorrectly triggers a firewall rule and causes an internal server error on several pages, including the author's profile page. (updated 2017-03-30)
• collectd metric paths can be truncated, which causes multiple write attempts to the same file for different metrics. (updated 2017-07-10)
• After changing the visibility of a repository, wiki search results have a conflicting number of displayed search results. Administrators can reindex the wiki through the site admin dashboard. (updated 2017-11-09)
• The create team API endpoint returns a 500 error if LDAP Sync is enabled and the team already exists. (updated 2018-01-09)

Thanks!

The GitHub Team