2018-07-17 11:43 GMT-03:00 Fabio Pardi <f.pardi@xxxxxxxxxxxx>: > > > On 07/17/2018 04:05 PM, Neto pr wrote: >> 2018-07-17 10:55 GMT-03:00 Fabio Pardi <f.pardi@xxxxxxxxxxxx>: > >>> Also i think it makes not much sense testing on RAID 0. I would start >>> performing tests on a single disk, bypassing RAID (or, as mentioned, at >>> least disabling cache). >>> >> >> But in my case, both the 2 SSDs and the 2 HDDs are in RAID ZERO. >> This way it would not be a valid test ? Because the 2 environments are >> in RAID ZERO. >> >> > > in theory, probably yes and maybe not. > In RAID 0, data is (usually) striped in a round robin fashion, so you > should rely on the fact that, in average, data is spread 50% on each > disk. For the sake of knowledge, you can check what your RAID controller > is actually using as algorithm to spread data over RAID 0. > > But you might be in an unlucky case in which more data is on one disk > than in another. > Unlucky or created by the events, like you deleted the records which are > on disk 0 and you only are querying those on disk 1, for instance. > > The fact is, that more complexity you add to your test, the less the > results will be closer to your expectations. > > Since you are testing disks, and not RAID, i would start empirically and > perform the test straight on 1 disk. > A simple test, like dd i mentioned here above. > If dd, or other more tailored tests on disks show that SSD is way slow, > then you can focus on tuning your disk. or trashing it :) > > When you are satisfied with your results, you can build up complexity > from the reliable/consolidated level you reached. > > As side note: why to run a test on a setup you can never use on production? > > regards, > > fabio pardi > Fabio, I understood and I agree with you about testing without RAID, this way it would be easier to avoid problems unrelated to my test on disks (SSD and HDD). Can you just explain why you said it below? "As side note: why to run a test on a setup you can never use on production?" You think that a RAID ZERO configuration for a DBMS is little used? Which one do you think would be good? I accept suggestions because I am in the middle of a work for my research of the postgraduate course and I can change the environment to something that is more useful and really used in real production environments. Best Regards []`s Neto >
