Containers migrated to BTRFS + SSD

[drobbins]

Hi All,

As we move to our third-generation infrastructure, I am currently migrating containers to BTRFS+SSD (from ZFS and HDD+Optane) which may result in up to an hour of downtime, especially if you have a lot of data stored in your containers. But the result will be much, much, much faster IO. So if you have seen your uptime on your container change recently, this is why.

Best,

Daniel

[tux]

Hi,

I see two small issues since the migration :

1. /usr/srv/linux symlink was broken on my container, ln -s host/linux linux fixed it.
2. disk usage is not related to container anymore, do you know if we can monitor it ? It helps keep instances lower on disk usage :-)

# df -hP
Filesystem      Size  Used Avail Use% Mounted on
/dev/loop0      600G  246G  352G  42% /
none            492K  492K     0 100% /dev
udev             10M     0   10M   0% /dev/tty
tmpfs           100K     0  100K   0% /dev/lxd
/dev/sda3       822G  592G  230G  73% /usr/src/host
tmpfs           100K     0  100K   0% /dev/.lxd-mounts
none             63G     0   63G   0% /dev/shm
tmpfs            13G  248K   13G   1% /run
cgroup_root      10M     0   10M   0% /sys/fs/cgroup

Best

Gael