> Your test dataset is too small and you aren't flushing the cache before > exiting dd, so you are largely seeing the time it takes to write to cache, > not to disk. > But that gives the RAID10 system 220 IOPs, still nowhere near the 100,000 > IOPs of a single SSD. > I suggest that you google a bit on how to do fileystem benchmarks first, then > try it and report back if something is still odd. > Your test dataset is too small and you aren't flushing the cache before > exiting dd, so you are largely seeing the time it takes to write to cache, > not to disk. . . . Oh, well, yes. I knew that I was "seeing" something that wasn't quite right. Your answers grounded me on such issues. Thank you und Entschuldigung! lbrtchx
On 6/17/20, Anders Andersson <pipat...@gmail.com> wrote: > On Wed, Jun 17, 2020 at 12:15 PM Albretch Mueller <lbrt...@gmail.com> > wrote: >> >> HDDs have their internal caching mechanism and I have heard that the >> Linux kernel uses RAM very effitiently, but to my understanding RAM >> being only 3-4 times faster doesn't make much sense, so I may be doing >> or understanding something not entirely right. > > I suggest that you google a bit on how to do fileystem benchmarks > first, then try it and report back if something is still odd. There > are many ways but "dd" is not the way unless you really dig through > the sync flags and understand what they do. I normally use "fio" but > it's not very friendly (so it suits me). > > However, I just recently put a fast NVMe SSD in an older server with > (lots) of DDR3 ECC RAM. The RAM bandwidth for one node/CPU is about > 10-12 GB/s, and the SSD bandwidth is nearing 2 GB/s for most loads. > That's getting close to your figures! > >
