0

I am buying a new server than has two HDs: 300 GB with 15000 rpm. 600 GB with 10000 rpm.

I am thinking to allocate the datafiles in both hard disks. and put the less used tables in the data files that reside in the slower HD that is 600 GB with 10000 rpm.

So, Can we create a table in specific data file - ?

I am using Oracle 11g R2 SOE .

Regards,

Yasir Arsanukayev
3,1653 gold badges23 silver badges30 bronze badges
asked Jun 13, 2013 at 7:50
3
  • 1
    You can only put tables in tablespaces, not in data files. Commented Jun 13, 2013 at 7:54
  • Have a read of the OFA documentation. Commented Jun 13, 2013 at 8:49
  • Arguably, for very frequently used and cached tables you might put them on the slower disks. In any case I'd be sure to look at v$_segment_stats for physical i/o when deciding which to place where. Commented Jun 13, 2013 at 10:27

1 Answer 1

4

You can create two tablespaces, put all the data files for one tablespace on one drive and all the data files for the other tablespace on the other side, and then create the tables in whichever tablespace you would like. It seems unlikely, however, that you would really want to do this. It is very unusual to run a database on a server without any sort of RAID configuration. And it is generally the case that any automatic I/O load balancing that you would implement would outperform your manual load balancing.

answered Jun 13, 2013 at 7:57
2
  • -1 for the last comment. While I am all to have the system do the work, it is not THAT Unusual to have DAS (Directly Atached Storage) and separate core high throughput and lower throughput for example for archive tables. I Know of some databases (one here in my work) deploying three levels actually - because some tables see a TON of activity, others do not, especially in the update area. A Raid controller will not magically make a decent decision - especially not taking financials into account. Commented Jun 13, 2013 at 8:19
  • 3
    @TomTom - I did say "generally". There are certainly systems that deal with a lot of historical data where having storage tiers makes sense. Though those systems would still have a separate RAID array for each storage tier rather than using a single drive. It is still generally the case that it would be very difficult for most systems to figure out how to manually distribute objects to maximize performance in this sort of setup unless you're willing to spend ridiculous amounts of time monitoring and redistributing load. Commented Jun 13, 2013 at 9:18

Your Answer

Draft saved
Draft discarded

Sign up or log in

Sign up using Google
Sign up using Email and Password

Post as a guest

Required, but never shown

Post as a guest

Required, but never shown

By clicking "Post Your Answer", you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.