NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
Striping granularity of RAID0
Merry x-mas, dear community!
On my RN104 guineapig I have created a RAID0 array and would like to know what chunk size determines the striping granularity. Researching turns out, that "normally...
Merry Christmas!
Retired_Member wrote:
On my RN104 guineapig I have created a RAID0 array and would like to know what chunk size determines the striping granularity.
There are a couple of ways to find this:
- It's in the log zip file (mdstat.log).
- # cat //proc/mdstat
- # mdadm --detail /dev/md127
It is set to 64K for RAID-5, so it likely is the same for RAID-0. Let us know what you find out.
Retired_Member wrote:
Any ideas, why it switched to RAID4 during the process?
Perhaps it just (falsely) reported that for some reason??? The first 4 disks of a 5 disk RAID4 array would have the same layout as your 4 disk RAID0 array.
Your assumption about the array, StephenB, is not completely true. For performance comparison reasons in this nas I have two volumes, one RAID0 holding 2 disks in the first two bays from the left and one RAID1 with two disks in the remaining bays.
However, I also thought, that it is a wrong comment, as RAID4 to my knowledge consists of minimum 3 disks with one disk dedicated to hold the parity data. With two disks in a RAID0 it would physically not be possible to switch to RAID4, though.
Retired_Member wrote:
RAID4 to my knowledge consists of minimum 3 disks with one disk dedicated to hold the parity data. With two disks in a RAID0 it would physically not be possible to switch to RAID4, though.
Yes, and that is the mode used in the v1 products.
All I was intending to point out was that (ignoring the dedicated parity disk), the data disks for RAID0 and RAID4 are identical. So is it possible to report a RAID0 array as a degraded RAID4 array (with a missing parity disk) and I was speculating that mdadm was doing that.
Back on point, are you seeing any difference in performance with the bigger chunk size?
StephenB wrote "Back on point, are you seeing any difference in performance with the bigger chunk size?"
No, not so far. Did not do any testing so far. Also fear, that the RN104's limited resources might lead to wrong conclusions as sometimes I get the impression the bottleneck is processor and/or memory and not disk capabilities and/or network.
Anyway, to know, that rechunking can be done on a RAID0 (or in general) is a great result. I might consider to do that under different circumstances on a different nas.
Thanks again, StephenB, your input was (as usual) very helpful and much appreciated. Simply said, I could not have done that alone :-)
Retired_Member wrote:
the RN104's limited resources might lead to wrong conclusions as sometimes I get the impression the bottleneck is processor and/or memory and not disk capabilities and/or network.
That is certainly true. I get about 70 MB/s read speeds on the RN102, and get 100 MB/s with the same model disks in an RN202.
NETGEAR Academy
Boost your skills with the Netgear Academy - Get trained, certified and stay ahead with the latest Netgear technology!
Join Us!
