While working on our flatpak build infrastructure I noticed that the flatpak is much larger than it should be. After quickly digging into the flatpak one of the culprits was quickly found; a bunch of unnecessary files such as include files were left.
Incident Report: Backend Down
Earlier this morning, at 04:38 UTC, one out of the twenty-two IMAP backends in production stopped serving its mail spool, showing Input/Output errors on its disk. Our Standard Operating Procedure is to examine log files, flush vm caches, stop the virtual machine, and start it back up again. This occurred at 05:48 UTC. The IMAP backend in question did not come back up cleanly.
Junk Email Filter.com is Junk
We’re dropping our use of junkemailfilter.com “Spam DNS Lists”, because we have few positive experiences with it. Frankly, it is Junk.
Update on the Kube flatpak nvidia driver issue
After some investigation into the nvidia driver issue, the flatpak has now been rebuilt based on the org.freedesktop.Platform runtime.
A limitation of flatpak currently keeps flatpak from making the appropriate driver available from any other runtime.
As a nice side-effect the overall size of the flatpak has been reduced by removing some unnecessary parts.
If you have already installed Kube, please follow the following instructions to update to the latest version (in all other cases, install it now):
Update to the latest version (please note this will also automatically install the new runtime and driver extensions):
flatpak -y --user update com.kolabnow.kube
To remove the no longer relevant platform, run the following commands:
flatpak --user uninstall org.kolab.Sdk
flatpak --user uninstall org.kolab.Platform
To check if the relevant driver was installed you can use the following command:
flatpak --gl-drivers
If you have an nvidia card, this should give you a line similar to (depending on your nvidia driver):
nvidia-384-90
To run kube, use:
flatpak run com.kolabnow.kube
Please let us know in the comments if this resolves the issue for you, thanks!
For more info about Kube, please head over to About Kube.
Announcing Service Windows: Implementing 2FA
As we recently announced, we have pursued an opt-in second factor authentication feature on Kolab Now. As described, the implementation is limiting users to the web client, and this requires some reconfiguration of various servers and services.
Kube flatpak not working with nvidia drivers
It was recently pointed out that the Kube flatpak for Kolab Now does not work with nvidia drivers, leaving you with error messages like this:
libGL error: No matching fbConfigs or visuals found libGL error: failed to load driver: swrast Unrecognized OpenGL version Unrecognized OpenGL version
As it turns out this is a problem with the extension system of flatpak that is used to provide the appropriate graphics driver to the application inside the container. It’s not quite clear what the best solution to the problem will be, but we’re working on it.
For more info about Kube, please head over to About Kube.
Pushing Support to our Enterprise Customer Support Platform
We are preparing a transition from our current platform underpinning the support@kolabnow.com email address — based on OTRS — to a more up-to-date, feature-rich environment based on Phabricator.
For various reasons, we much anticipate this change. Not in the least we’ll have real-time notifications about new tickets, and user responses, and a few chat rooms — it’ll allow us to sort our workload better.
TOTP-based Two Factor Authentication Passed QA
In a previous blog post, I have told you about our experimenting with TOTP-based two factor authentication. It proves functional in the Cockpit and in the Web Client, so we’re preparing the promotion to production.
A Stricter DMARC Policy, Part II
Last month, we let you know a stricter DMARC policy was being applied to Kolab Now infrastructure. With a primary aim to increase our reputation and decrease phishing attempts from clearly false senders, we’ve since learned about some secondary effects;
I Apologize for the Delay
If you’ve noticed our responses to support tickets or monitoring alerts is a little slower than usual, that’s because this is now the view from our office:
