RE: db_file_multiblock_read_count causing full scans to takelonger?

After more testing with dd on 2 HPUX 11 boxes, an AIX 5.2 and an AIX 5.3 box, here is what I found:

Both of the HPUX boxes showed a consistent drop in performance when going from bs=255k to bs=256k, but on average over several samples, it was usually only an increase of about 25-35% in real time (not 2x and higher like you are seeeing), e.g. from 15 seconds to 20 seconds, and as I increased the block size further up to 1024k, the performance improved and time dropped to about the same as I was getting at 255k.

I did not see this drop in performance at any level with AIX - it stayed fairly consistent from about 32k all the way up to 8192k.

So, there appears to be some overhead with HPUX that kicks in right at 256k, but it seems to get cancelled out by some other increased efficiency as the read size approaches 1M. Obviously this is just a generalization based on 4 machines so everybody don't run out and adjust your dbfmbrc based on this - YMMV. Curious if others see the same behavior with their HPUX filesystems too?

Privileged/Confidential Information may be contained in this message or attachments hereto. Please advise immediately if you or your employer do not consent to Internet email for messages of this kind. Opinions, conclusions and other information in this message that do not relate to the official business of this company shall be understood as neither given nor endorsed by it.

--
http://www.freelists.org/webpage/oracle-l