Dont know if this has been covered on here, or its a known issue, so I thought I would post it.
I have a Nethserver version 6.9 that has been running great in a production environment for a business since late 2015.
Lately the updating has been failing. Turns out the /boot partition was filling with old kernel versions. This can be checked by running the df command from the terminal. I have seen this before on ubuntu server, but have no other experience with CentOS.
Googling revealed that running yum update checks for and installs updates (and I assume that is what Nethserver does). But instead, you can run yum upgrade which removes all but the 3 most current kernels.
But that still failed. It said it needed 9mb of free space on /boot. So i cd’d into the /boot directory and manually deleted a few of the oldest kernels (be careful here). Then I ran yum upgrade again. It did give warnings about those deleted files missing, but it uninstalled all those old kernels and freed up a heap of space.
So, problem solved, until it happens next time.
So hopefully this helps someone out.
So I have 2 questions for the Nethserver team:
In Nethserver 7.4 and beyond, is this still an issue? I am no Linux expert, but I had this continually happening on ubuntu server, and found the solution was to do away with the /boot partition, and have /boot within the main / partition. And later versions of ubuntu server seemed to have done that.
Is there a way to prevent this from happening on Nethserver 6.9?
Thanks for the excellent software.