if the SST process was failed in the middle, it left temporary files inside SST temporary directory.

Kubernetes restarts Pod after failure and PXC starts SST again, but sometimes it is not enough free space for both copies on the disk because files from the previous SST still exist on disk.