Skip to main content

You are not logged in. Your edit will be placed in a queue until it is peer reviewed.

We welcome edits that make the post easier to understand and more valuable for readers. Because community members review edits, please try to make the post substantially better than how you found it, for example, by fixing grammar or adding additional resources and hyperlinks.

5
  • thanx for replying. i forgot to mention that i did call --examine on the device and nothing was printed about chunk-size. here's an example output. is it my mdadm version (v3.1.4) or metadata version that makes the difference?
    – Huang Tao
    Commented Aug 30, 2011 at 8:38
  • 1
    I've seen two thoughts on this. One, since RAID 1 doesn't stripe, chunk size is irrelevant. In another, it's only relevant to reads/seeks. My guess is, it wasn't set since it's RAID one, and if it has a value, it's the default 64k
    – OldWolf
    Commented Aug 31, 2011 at 16:03
  • Your output says you're using RAID1. See the comment and two links I've added above.
    – OldWolf
    Commented Aug 31, 2011 at 16:13
  • 3
    i'm aware that i was using the 64k default chunk-size. i've read the man pages and the wikipedia article before i asked the question. however, since "the chunk-size specifies how much data to read serially from the participating disks", it does matter for raid 1 arrays, doesn't it? i mean, it's supposed to take chunk-size into consideration tuning the filesystem lying on the array, right?
    – Huang Tao
    Commented Aug 31, 2011 at 17:05
  • You'd probably want the chunk size to be small enough so that application requests get distributed over drives and big enough to not spend a lot of time splitting or assembling data streams.For SSDs, it's get a bit more complicated as the internal erase block size of the drive will impact performance. A chunk size of 1 MiB or more can be a good choice. Commented May 12 at 12:19