On Mon, Mar 26, 2012 at 2:10 AM, zfs user <zf...@itsbeen.sent.com> wrote:
> On 3/25/12 10:25 AM, Aubrey Li wrote:
>> On Mon, Mar 26, 2012 at 12:48 AM, Richard Elling
>> <richard.ell...@richardelling.com> wrote:
>>> This is the wrong forum for general purpose performance tuning. So I
>>> continue this much farther. Notice the huge number of icsw, that is a
>>> symptom than locks.
>>> -- richard
>> thanks anyway, lock must be a problem. the scenario here is, apache
>> causes a bunch of stat()
>> syscall, which leads to a bunch of zfs vnode access, byond the normal
>> read/write operation.
>> The problem is, every zfs vnode access need the **same zfs root**
>> lock. When the number of
>> httpd processes and the corresponding kernel threads becomes large,
>> this root lock contention
>> becomes horrible. This situation does not occurs on linux.
>> Let me see if any others have a clue for me. Any suggestions will be
>> highly appreciated!
> Is noatime turned on for all your zfs filesystems?
> Have you tried the latest version of httpd?
yeah, atime is turned off. It's not helpful to reduce the zfs vnode access.
I didn't try the latest version of httpd. I want to get the same performance
by the same user land software stack as it on Linux.
zfs-discuss mailing list