I have a customer running R80.40 VSX clusters on 7k appliances (about 25 VS).
lv_current on one of the members was exhausted and they couldn't push a change. They found some old files they could delete, enough to claw back about 2GB so the changes could be applied.
The root partition has been gradually filling since install and TAC advise there are no other files that can be deleted to recover space - they recommend extending lv_current by 15-20GB which will also increase the upgrade volume and eat into available snapshot space.
While LVM is familiar to engineers, I'm trying to understand why this level of maintenance should be necessary on a device sold as an appliance. Wouldn't it make more sense to size the physical disk and volumes sufficient for normal operations for the life of the appliance, or automate disk optimisation like my TiVo did in 1999?
Disk space (or lack of it) appears to be a common source of frustration here.
