The difference between bigrow and the other tools is that bigrow is pre-allocating and re-formatting disk space specified by the command. If there is a difference in performance between disks, it is because one disk has a better I/O then the other. A test outside of Progress can be performed to eliminate Progress involvement.
For example, if specifying 8 in the bigrow command and the BI cluster size is 16Mb, that means the BI file is being pre-formatted to (8+4) * 16Mb = 192Mb. The 4 represents the default BI clusters. On unix/linux run the dd command to simulate what bigrow does:
/usr/bin/time dd if=/dev/zero of=/tmp/test.out bs=<bi block size> count=<value> oflag=dsync
The BI block size can be acquired from the output of the command "proutil <dbname> -C describe"
For the example above, the count should be a number that multiplied by the BI block size will get a file that is 192Mb in size, for example if the BI block size is 8k then the command should be :
/usr/bin/time dd if=/dev/zero of=./test.out bs=8k count=24550 oflag=dsync |
