Issues on Flathub Status

[Resolved] Publishing delay due to delta regeneration

Wed, 25 Mar 2026 00:58:00 +0000

We are in the process of regenerating deltas for the current apps. The process takes an exclusive lock of the underlying ostree repository, preventing new builds from being published. If you have an urgent update to release, please contact us on our Matrix channel. (02:14 UTC — Mar 25)

[Resolved] appstream2 branch referencing stale commit

Tue, 24 Mar 2026 23:55:00 +0000

A mismatch between the appstream2 branch and delta indexes caused local Flatpak clients to fail to update appstream data. The issue is resolved now after regenerating both the branch and the indexes.

If you are still experiencing issues, please try the steps below to clear the local copy of the appstream2 branch. (01:33 UTC — Mar 25)

# For system-wide installation
sudo ostree --repo=/var/lib/flatpak/repo refs --delete flathub:appstream2/x86_64
# For user installation
ostree --repo=$HOME/.local/share/flatpak/repo refs --delete flathub:appstream2/x86_64
# Update the appstream data
flatpak update --appstream

Additionally, some client experienced timeouts when fetching delta indexes that were removed from the server. This was caused by the CDN not caching 404 responses, resulting in requests queuing up behind nginx’s cache lock. A fix was deployed to cache 404 responses, allowing clients to gracefully fall back to full downloads. (02:08 UTC — Mar 25)

[Resolved] Signing outage

Tue, 17 Feb 2026 13:56:00 +0000

An issue with hardware security key is currently preventing builds from being signed with our GPG key. This affects both direct uploads and builds performed on Flathub infra. Since no builds can be signed thus published, the build service including test builds are also disrupted. (13:56 UTC — Feb 17)

We are in contact with the data center staff to investigate the issue physically. (17:49 UTC — Feb 19)

Data centre staff attempted reseating the signing key across available USB ports without success. We are now arranging delivery and installation of a replacement key. (16:46 UTC — Feb 20)

We successfully managed to install the new Yubikey. (14:04 UTC — Feb 23)

Information

We will soon restart all affected official builds in batches and they should be published within a couple of hours. Affected direct uploaders will be informed to restart builds from their end. Once this process is complete, some services which were disabled (to reduce impact) will be re-enabled one by one.

Once the test build service is available, any affected test builds on pull requests will need to be manually restarted by commenting “bot, build” by the respective maintainer or contributor.

Please allow us some time to finish the entire process outlined above. Publishing may take longer than usual due the large number of builds in the queue. Please open an issue on GitHub for any further query about missing or unpublished builds. Further updates will be posted in the statuspage.

All affected builds were successfully restarted and they finished building. (16:40 UTC — Feb 23)

The repository server is still experiencing some issues. (17:30 UTC — Feb 23)

The incident should be resolved now and services are being restarted. (18:50 UTC — Feb 23)

[Resolved] CDN cache invalidation issue

Fri, 13 Feb 2026 05:38:00 +0000

Due to a brief downtime, one of the origin servers missed purging operation and was serving stale and non-existent index files. As a result Flathub builds failed while attempting to download dependencies and any general Flatpak clients experienced hangs or download failures.

The issue was resolved after the cache was invalidated. (07:00 UTC — Feb 14)

[Resolved] CDN cache invalidation issue

Tue, 13 Jan 2026 05:51:00 +0000

Due to a CDN cache invalidation issue, the CDN served a summary file referencing a non-existent file, causing Flatpak clients to hang or fail with 404 errors. (05:51 UTC — Jan 13)

Resolved: The issue was resolved after invalidating the cache again. (06:40 UTC — Jan 13)

[Resolved] Website facing intermittent disruptions

Sat, 13 Sep 2025 15:29:13 +0000

The Flathub website frontend is currently undergoing some rework, and users may experience intermittent disruptions while accessing it. We are trying to address this. (15:29 UTC — Sep 13)

Resolved: This should be resolved now (08:44 UTC — Sep 17)

[Resolved] Publish outage

Tue, 27 May 2025 11:47:00 +0000

An ongoing outage is currently preventing builds from being published and is causing test build links to return 404 errors. This issue affects both direct uploads and all builds performed on Flathub infra.

We are trying to resolve it. (11:47 UTC — May 27)

This is resolved now. (10:04 UTC — May 28)

[Resolved] Build service outage

Sun, 03 Nov 2024 19:05:00 +0000

There is an ongoing outage that is causing builds to fail at the commit stage. This affects all builds done through Flathub buildbot and may affect direct uploads as well.

We are trying to resolve it. (19:05 UTC — Nov 3)

Resolved: This should be resolved now (19:39 UTC — Nov 4)

Hosting provider planned maintenance

Mon, 07 Oct 2024 18:00:00 +0000

Our hosting provider will be undergoing network maintenance starting at 6:00 UTC omn October 8th, 2024. The website, Discourse and build service will be unavailable during this time. The CDN and main repository server will remain available. This does not affect the current status.

[Resolved] Blocked publishing due to ongoing repository pruning

Tue, 14 May 2024 10:00:00 +0000

Resolved: The pruning has completed, freeing almost 7TB of disk space. Affected builds have been restarted in Buildbot. (17:50 UTC — May 15)

The pruning process is still ongoing. It may take about 12 hours more. (05:30 UTC — May 14)

We are in the process of pruning the ostree repository from old builds. The process requires an exclusive lock of the repository, meaning that publishing is not going to be possible until it finishes. We will manually publish the outstanding builds afterwards. (19:00 UTC — May 14)

[Resolved] Frontend service is down

Tue, 23 Apr 2024 17:10:00 +0000

Resolved: While we have identified problematic code, we are going to keep the previous version running until tomorrow. (20:20 UTC — Apr 23)

The service has been rolled back to a previous version. The frontend service is now operational. We are still investigating a proper fix. (19:59 UTC — Apr 23)

The frontend service is unable to serve requests. We’re investigating recent code changes that could have caused this. (17:50 UTC — Apr 23)