For some reason one physical server and two virtual servers would not boot after the new kernel was installed. Something went wrong during the install process and messed up the grub configuration on two of the machines, I was able to repair these with the boot-repair disk from sourceforge. But a third machine was so scribbled I had to restore it from backups.
At this point everything is back in service. I apologize for the longer than anticipated outage.
One physical server that hosts among other things the mail spool and server failed to boot. I have to drive down to the co-lo to reboot it so service is going to take longer than expected to restore.
We were forced to downgrade MariaDB back to 10.5 because 10.6.1 deprecated some functionality required by NextCloud. I have filed a bug report with Nextcloud so they can be aware their code is out of date and hopefully fix it.
I’ve got about half a dozen of the major sources blocked and while there is still some packets dribbling through from others they are not enough to cause network or router overload so should not further impact network performance.
The kernel upgrades are completed and all NFS and NIS connections have been verified. The upgrade was relatively uneventful. Everything is back up.
I am in the process of upgrading vps3 because the version of Devuan has reached end of life and it no longer has repositories from which to receive security updates. At the conclusion of this process there will be about 1/2 hour downtime for it to image it.
The free public services https://friendica.eskimo.com/, https://hubzilla.eskimo.com/, and https://nextcloud.eskimo.com/ are back in service.
If you get e-mail like this:
And your user@eskimo.com is still using old version, Please tap the blue button below to upgrade your mailbox to the latest version, and get 25GB free data space with a more organized mailbox to avoid deactivation.
It is a scam. You can forward it to spamtrap@eskimo.com or just delete it. I’ve already traced the source, it is coming from a Seattle based competitor. Slimy business practice in my view but I guess that’s our world today.
I hope to do kernel upgrades tonight from 5.12.4 to 5.12.8, mostly a bug fix upgrade. Because there is no change in major point release, hopefully it will go smoothly.
This will result in approximately 10 minute outages of various services during this time frame, this includes shell servers, virtual private servers, web hosting, email, and https://friendica.eskimo.com/, https://hubzilla.eskimo.com/, and https://nextcloud.eskimo.com/.
I’m going to take vps3, vps5, and vps7 down after 11PM and before 2AM for maintenance (to image, a form of backup) for about 1/2 hour each tonight.
Maintenance on Fedora, Centos7, Scientific7, and JuLinux has been completed.
All of these machines will be going down for 15-45 minutes this afternoon to image (a form of backup). They will be taken down one at a time and I will try to avoid / delay those actively in use.