Either true Index Organized Tables or Clustered Indexes would be very useful for a variety of table/query types. The latter seems more difficult with Postgres' MVCC model since it requires data to be stored in the index that is authoritative. Storing the full tuple in an index and not even having a data only page would also be an interesting approach to this (and perhaps simpler than a separate index file and data file if trying to keep the data in the order of the index). On 7/15/09 7:17 PM, "Justin Pitts" <justinpitts@xxxxxxxxx> wrote: > Is there any interest in adding that (continual/automatic cluster > order maintenance) to a future release? > > On Wed, Jul 15, 2009 at 8:33 PM, Scott Carey<scott@xxxxxxxxxxxxxxxxx> wrote: >> If you have a lot of insert/update/delete activity on a table fillfactor can >> help. >> >> I don¹t believe that postgres will try and maintain the table in the cluster >> order however. >> >> >> On 7/15/09 8:04 AM, "Ibrahim Harrani" <ibrahim.harrani@xxxxxxxxx> wrote: >> >> Hi, >> >> thanks for your suggestion. >> Is there any benefit of setting fillfactor to 70 or 80 on this table? >> >> >> >> On Wed, Jun 24, 2009 at 8:42 PM, Scott Marlowe<scott.marlowe@xxxxxxxxx> >> wrote: >>> As another poster pointed out, you cluster on ONE index and one index >>> only. However, you can cluster on a multi-column index. >>> >> >> -- >> Sent via pgsql-performance mailing list (pgsql-performance@xxxxxxxxxxxxxx) >> To make changes to your subscription: >> http://www.postgresql.org/mailpref/pgsql-performance >> >> > -- Sent via pgsql-performance mailing list (pgsql-performance@xxxxxxxxxxxxxx) To make changes to your subscription: http://www.postgresql.org/mailpref/pgsql-performance
