[email protected] wrote: > Squashfs is also an overhead because it decompress the contents of fs. > Generally every overhead has its pros and cons. In squashfs, you pay the > cpu cycles for decompression, but you get more disk space, for instance.
Not completely related to aufs but we did a fair amount of real world benchmarking with aufs/squashfs combos. One handy tool is measuring boot times using bootchart which shows cpu and IO utilization. In general we found that squashfs improved performance by quite a bit since the cpu can decompress data much faster than it can be read from a hard disk. With flash drives it was a little more complicated and depended on speed. With faster flash drives and slower cpus we could actually observe the machine becoming CPU bound rather than IO bound. In those cases, it's actually more efficient to leave the data uncompressed. Ultimately, I think it's important to gather data about the particular hardware you're using before deciding how to lay out your data. cheers, Kris ------------------------------------------------------------------------------ Join us December 9, 2009 for the Red Hat Virtual Experience, a free event focused on virtualization and cloud computing. Attend in-depth sessions from your desk. Your couch. Anywhere. http://p.sf.net/sfu/redhat-sfdev2dev
