Reputation: 884
I have two table phone and phone_area
phone columns:
id: pk
phone: unique index
phone data:
|id| phone |
+--+-------------+
|1 | 1882601xxxx |
+--+-------------+
|2 | 1882602xxxx |
+--+-------------+
|2 | 1882602xxxx |
+--+-------------+
|2 | 1882603xxxx |
+--+-------------+
phone_area columns:
1.id: pk
2.phone: unique index
3.area: varchar(20)
phone_area data:
|id| phone | area |
+--+--------+----------+
|1 | 1882601| area_one|
+--+--------+----------+
|2 | 1882602| area_two|
+--+--------+----------+
|2 | 1882603|area_three|
+--+--------+----------+
My sql following:
SELECT t1.phone,t2.area FROM phone t1
LEFT JOIN phone_area t2 ON substr(t1.phone, 1, 7)= t2.phone
It is very slowly.
When explain the sql, it shows the type as "ALL" and Using where; Using join buffer (Block Nested Loop)
How to improve my sql?
Upvotes: 1
Views: 83
Reputation: 40481
Follow this steps:
ALTER TABLE phone ADD phone_cut int; --add a new column
UPDATE phone SET phone_cut = substr(phone, 1, 7); --store the cropped value of phone
ALTER TABLE phone ADD INDEX ind_name (phone_cut); -- add an index on that column
And then :
SELECT t1.phone,t2.area FROM phone t1
LEFT JOIN phone_area t2
ON t1.phone_cut= t2.phone
Upvotes: 2