RE: sreadtim and mreadtim < 1ms

From: Allen, Brandon <>
Date: Tue, 3 Nov 2009 14:27:16 -0700
Message-ID: <64BAF54438380142A0BF94A23224A31E112E6AF92C_at_ONEWS06.oneneck.corp>

There seems to be some confusion about what I was asking in my original post - I already understand why I'm getting such fast sreadtim and mreadtim values (from OS caching). I'm just trying to figure out what is the best approach (in general) for setting the system stats in this scenario:
  1. Leaving them unset and using the default stats (ioseektim=10, iotfrspeed=4096) - this is what I have currently in place
  2. Gathering noworkload stats (I get ioseektim=5 and iotfrspeed=35632)
  3. Leaving them at their very small values as gathered by gather_system_stats (sreadtim=.1, mreadtim=.3)
  4. Manually setting them higher so the CBO won't ignore them, e.g. sreadtim=1, mreadtim=3
  5. Other options I haven't considered?

I know testing every single one of my application's queries with each different setting is the only way to know for sure, but I'm just looking for any general guidelines, rules of thumb, best practices, personal experiences, etc. that others may have had in this same situation. I figure there must be lots of people in the same situation with the price & size of RAM these days. I wish I had 11g with the Real Application Testing license so I could run Database Capture & Replay with the different stats in place, but this is only 10g Standard Edition.


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.
Received on Tue Nov 03 2009 - 15:27:16 CST

Original text of this message