| List: | General Discussion | « Previous MessageNext Message » | |
| From: | Thimble Smith | Date: | July 23 1999 2:24am |
| Subject: | Re: Why this index? | ||
| View as plain text | |||
At 19:11, 19990722, Steven Roussey wrote: >170 (63 with c=0) rows match 15950 and 19687 (3586 with c=0) rows match >12750. > >With so many rows matching, it seems to me that it ought to use the 8 byte >index(a,c) that matches the where clause rather than the first 4 bytes of >the index(a,b). I could be way off, but have you used isamchk -a to inform MySQL about your tables? It sometimes does real wonders for the proper selection of keys to use. Tim
| Thread | ||
|---|---|---|
| • Why this index? | Steven Roussey | 23 Jul |
| • Re: Why this index? | Paul DuBois | 23 Jul |
| • RE: Why this index? | Steven Roussey | 23 Jul |
| • Re: Why this index? | Thimble Smith | 23 Jul |
| • RE: Why this index? | Steven Roussey | 23 Jul |
| • RE: Why this index? | Michael Widenius | 2 Aug |
| • Schema Generator | Bradley Glonka | 23 Jul |
| • Re: Schema Generator | Micheal Mc Evoy | 23 Jul |
| • Re: Schema Generator | Sasha Pachev | 23 Jul |
| • Why this index? | sinisa | 23 Jul |
| • RE: Why this index? | Steven Roussey | 23 Jul |
| • Why this index? | Michael Widenius | 2 Aug |
