[MDEV-7239] explain index choice Created: 2014-12-01 Updated: 2019-06-14 Resolved: 2019-06-14 |
|
| Status: | Closed |
| Project: | MariaDB Server |
| Component/s: | Optimizer |
| Fix Version/s: | N/A |
| Type: | Task | Priority: | Minor |
| Reporter: | roberto spadim | Assignee: | Sergei Petrunia |
| Resolution: | Duplicate | Votes: | 1 |
| Labels: | None | ||
| Issue Links: |
|
||||||||
| Description |
|
Hi guys, i openned this MDEV from
what's the idea of this mdev? my idea is include one new column with a dynamic output (json or dynamic column)
i don't know the complexity of this mdev but it's interesting to better know what optimizer do |
| Comments |
| Comment by Elena Stepanova [ 2014-12-01 ] |
|
Please don't create multiple entries about the same problem, it serves no purpose at all, only makes things more complicated. If you want to share your thoughts or confusion, or have questions, send it to the mailing list. If you want to provide more information on the existing issue, comment on it. |
| Comment by Elena Stepanova [ 2014-12-01 ] |
|
Closing as a duplicate of |
| Comment by roberto spadim [ 2014-12-01 ] |
|
hum, this is not a duplicate of mdev-7125, it's a feature (task) about extending explain command not a bug report about optimizer choise |
| Comment by Elena Stepanova [ 2014-12-01 ] |
|
Sorry, misunderstood. Re-opened. |
| Comment by roberto spadim [ 2014-12-01 ] |
|
=) no problem |
| Comment by roberto spadim [ 2014-12-01 ] |
|
the only 'information' that explain i don't know the optimizer choise, but what i see is 164003 < 226330, maybe this explain the optimizer choise? |
| Comment by Sergei Petrunia [ 2014-12-02 ] |
|
MySQL 5.6+ has optimizer_trace. The goal of optimizer_trace was exactly this - to provide clues how the optimizer processed the query - what options were considered, and why some of them were discarded. We haven't backported optimizer_trace to MariaDB because some of MariaDB had concerns about whether optimizer_trace is a finished work (it has certain issues). Recently we've been kind of leaning towards backporting optimizer trace in its current form. At the moment, I don't see any other solution to this problem except optimizer trace. |
| Comment by Sergei Petrunia [ 2014-12-02 ] |
|
Well, there is also the traditional manual way: try the query with IGNORE INDEX for the index that was used and see what explain plan you will get. |
| Comment by roberto spadim [ 2014-12-02 ] |
|
nice |
| Comment by Sergei Petrunia [ 2019-06-14 ] |
|
Marking as duplicate of |