On 14/10/16 14:31, Mark Hahn wrote:
anyway if I force direct I/O, for example using oflag=direct in dd, the write performance drop as low as 8MB/sec

with 1MB block size. And each write it's about 120ms latency.

but that's quite a small block size. do you approach buffered performance if you write significantly bigger blocks (8-32M)? presumably you're already
striping across OSTs?
Yes I agree with you but my lustre filesystem can have many small files so I Was trying to do a test similar
to the real life situation.
lustre-discuss mailing list

Reply via email to