Connor - so you are saying come up with a baseline then actually set your stats to that for those tables and indexes then periodically check those stats? =20 -----Original Message----- From: oracle-l-bounce@xxxxxxxxxxxxx [mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Connor McDonald Sent: Tuesday, April 12, 2005 8:40 AM To: 'oracle-l@xxxxxxxxxxxxx' Subject: Re: VLDBA's: gather stats on a large table Couple of ideas... a) all the stats gathering facilities do a *lot* of sorting. See if you can drop your 20hours by tinkering with the various pga parameters and/or parallel. b) but mostly when it comes to bigger tables, I'm a fan of "when in doubt, cheat"...If *you* have got a good idea of the stats, just use set_..._stats. If you don't have a decent starting point, sacrifice (say) 60 hours on a weekend to get the best stats you can, and then use them as base for subsequent "cheating". Every few months/years/etc, take some stats - see how they compare to your cheats and adjust accordingly. =20 hth connor --- "Teehan, Mark" <mark.teehan@xxxxxxxx> wrote: >=20 > I have a 7bn row, weekly partitioned IOT that adds hundreds of=20 > millions of rows each week. A five percent stats job estimate takes 20 hrs every weekend. > Given that I only insert to the latest partition, all others are RO,=20 > and have no global indexes, how should I gather stats? As the app=20 > doesnt specify the partition in selects, I need to maintain global=20 > stats, which means I need to rescan all 7bn rows each time. Is there a > better way to do this? How to you gather stats on data that is 90% read-only, but maintain accurate global stats? >=20 > Thanks! > Mark Teehan >=20 > = =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > =3D=3D=3D=3D=3D=3D=3D=3D This message is for the sole use of the = intended recipient.=20 > If you received this message in error please delete it and notify us.=20 > If this message was misdirected, CSFB does not waive any=20 > confidentiality or privilege. CSFB retains and monitors electronic=20 > communications sent through its network. Instructions transmitted over > this system are not binding on CSFB until they are confirmed by us. Message transmission is not guaranteed to be secure. > = =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D > =3D=3D=3D=3D=3D=3D=3D=3D >=20 > -- > //www.freelists.org/webpage/oracle-l >=20 Connor McDonald Co-author: "Mastering Oracle PL/SQL - Practical Solutions" Co-author: "Oracle Insight - Tales of the OakTable" web: http://www.oracledba.co.uk web: http://www.oaktable.net email: connor_mcdonald@xxxxxxxxx "GIVE a man a fish and he will eat for a day. But TEACH him how to fish, and...he will sit in a boat and drink beer all day" ------------------------------------------------------------ =09 __________________________________ Do you Yahoo!?=20 Yahoo! Small Business - Try our new resources site! http://smallbusiness.yahoo.com/resources/ -- //www.freelists.org/webpage/oracle-l BEGIN-ANTISPAM-VOTING-LINKS ------------------------------------------------------ Teach CanIt if this mail (ID 29381537) is spam: Spam: https://dohsmsi01.doh.state.fl.us/canit/b.php?c=3Ds&i=3D29381537&m=3D5587= 6defa ab3 Not spam: https://dohsmsi01.doh.state.fl.us/canit/b.php?c=3Dn&i=3D29381537&m=3D5587= 6defa ab3 Forget vote: https://dohsmsi01.doh.state.fl.us/canit/b.php?c=3Df&i=3D29381537&m=3D5587= 6defa ab3 ------------------------------------------------------ END-ANTISPAM-VOTING-LINKS -- //www.freelists.org/webpage/oracle-l