Question

在Memsql中执行以下查询时，收到以下消息：

“ MemSQL Distributed不支持功能'带有未在下推查询中分片的子选择的联合'。”

SELECT *
              FROM customer_campaigns t45 LEFT OUTER JOIN (SELECT
                                                             t164.source_entity_id      AS source_entity_id,
                                                             t164.destination_entity_id AS destination_entity_id,
                                                             t164.status                AS STATUS
                                                           FROM campaign_mirror_references t164 INNER JOIN
                                                             mirror_configurations t160
                                                               ON ((t164.mirror_config_id = t160.id) AND
                                                                   (t160.is_sync = 1))
                                                             INNER JOIN customer_campaigns t45
                                                               ON (t45.campaign_id = t160.source_campaign_id)
                                                           WHERE (t45.cust_profile_id = 416)
                                                           UNION DISTINCT SELECT
                                                                            t164.source_entity_id      AS source_entity_id,
                                                                            t164.destination_entity_id AS destination_entity_id,
                                                                            t164.status                AS STATUS
                                                                          FROM
                                                                            campaign_mirror_references t164 INNER JOIN
                                                                            mirror_configurations t160 ON (
                                                                            (t164.mirror_config_id = t160.id) AND
                                                                            (t160.is_sync = 1) AND
                                                                            (t160.dest_profile_id = 416))) AS SOURCE
                  ON ((t45.campaign_id = SOURCE.source_entity_id) AND (SOURCE.status <> 2))

Answer 1

是否要求将其编写为单个查询？为了提高可读性并简化嵌套，您可以尝试制作一个临时表然后从中进行选择吗？这样应该会获得与上面相同的效果，但是更容易理解：

DROP TABLE IF EXISTS tmpCampaignMirrorRefs;

CREATE TEMPORARY TABLE tmpCampaignMirrorRefs AS
SELECT t164.source_entity_id AS source_entity_id
    , t164.destination_entity_id AS destination_entity_id
    , t164.status AS STATUS
FROM campaign_mirror_references t164
INNER JOIN mirror_configurations t160 ON t164.mirror_config_id = t160.id
    AND t160.is_sync = 1
WHERE t45.cust_profile_id = 416
UNION
SELECT t164.source_entity_id AS source_entity_id
    , t164.destination_entity_id AS destination_entity_id
    , t164.status AS STATUS
FROM campaign_mirror_references t164
INNER JOIN mirror_configurations t160 ON t164.mirror_config_id = t160.id
    AND t160.is_sync = 1
    AND t160.dest_profile_id = 416
;

SELECT *
FROM customer_campaigns t45
LEFT OUTER JOIN tmpCampaignMirrorRefs SOURCE ON t45.campaign_id = SOURCE.source_entity_id
    AND SOURCE.status <> 2
;

让我知道是否还有其他问题-希望有帮助！

也是FWIW，在MemSQL中为UNION == UNION DISTINCT

MemSQL-如何在子查询中修复与子联合

1 个答案: