Good morning,
Some of you may have experienced a small hiccup this morning (eastern time). I went ahead and upgraded our lemmy instance to 0.17.4. I also took this opportunity to make some small adjustments to the instance that required a restart of services. These changes were mainly related to increasing Federation worker count in hopes to decrease synchronization times with other instances. A small fix to help slow down the storage growth was also applied (limiting log sizes that get generated).
Everything should be back up now. In the future my goal is to be more transparent with scheduling downtime.
edit: updated link
As I understand it, part of the issue on kbin is that the default federation threads value is too low for large sized instances.
I gather this will improve over time.