Cumulus Support

Whatever, harddisks are things they should have in store in abundancy.
Running to some store to buy a sixpack is , well... fill in yourself.

HansR wrote: ↑Thu 05 Sep 2024 2:53 pm Running to some store to buy a sixpack is , well... fill in yourself.

You clearly missed my sense of humour.

But disk failures wouldn't normally cause a loss of capacity, it would reduce redundancy, or if multiple fail together you have a complete loss of a storage system and you must recover it.

mcrossley wrote: ↑Thu 05 Sep 2024 4:07 pm But disk failures wouldn't normally don't cause a loss of capacity, it would reduce redundancy, or if multiple fail together you have a complete loss of a storage system and you must recover it.

I know. You’d think they’d configure them in a RAID configuration so even if one fails the others carry on.

They did say it was a case of running short of storage hence why uploads were blocked.

If their systems are even remotely modern, they are running things on virtual machines, and they also likely use shared storage and thin provisioned disks. In unfortunate events there are many things that can cause the allocated virtual disk space to be filled rapidly and unexpectedly. I’m not defending them since disk space mangement and monitoring on both virtual machine and server scope are essential features to be taken care of. But just saying it can happen, and for a virtual machine the allocated space can run out even if there is plenty of space available on the physical storage.