Since current D06 systems aren't bootable with the previous kernel, I restricted my regression testing to D05. I ran iozone -a on the current and -proposed kernels, then compared the results visually using iozone-results-comparator. While the changes in read performance was all negligible, there were improvements across the board with the write and re-write tests.
Verification:
Since current D06 systems aren't bootable with the previous kernel, I restricted my regression testing to D05. I ran iozone -a on the current and -proposed kernels, then compared the results visually using iozone- results- comparator. While the changes in read performance was all negligible, there were improvements across the board with the write and re-write tests.