我们在 Redshift 中有一个表:
人
people_id people_tele people_email role
1 8989898332 <a href="https://stackoverflow.com/cdn-cgi/l/email-protection" class="__cf_email__" data-cfemail="4a202522240a2d272b232664292527" rel="noreferrer noopener nofollow">[email protected]</a> manager
2 8989898333 <a href="https://stackoverflow.com/cdn-cgi/l/email-protection" class="__cf_email__" data-cfemail="27545342514267404a464e4b0944484a" rel="noreferrer noopener nofollow">[email protected]</a> manager
3 8989898334 <a href="https://stackoverflow.com/cdn-cgi/l/email-protection" class="__cf_email__" data-cfemail="93f2fdf7e1f6e4d3f4fef2faffbdf0fcfe" rel="noreferrer noopener nofollow">[email protected]</a> manager
4 8989898335 <a href="https://stackoverflow.com/cdn-cgi/l/email-protection" class="__cf_email__" data-cfemail="98fffdf7eafffdd8fff5f9f1f4b6fbf7f5" rel="noreferrer noopener nofollow">[email protected]</a> manager
我有一些用户会查询该表，例如:
select * from people where role = 'manager' limit 1;
系统用户基本上是调用这些人来推销产品。因此，当查询返回结果时，它不应该返回相同的人。
例如。
如果用户 A 执行查询 - select * from people where role = 'manager' limit 1;，那么他应该得到结果:
people_id people_tele people_email role
1 8989898332 <a href="https://stackoverflow.com/cdn-cgi/l/email-protection" class="__cf_email__" data-cfemail="96fcf9fef8d6f1fbf7fffab8f5f9fb" rel="noreferrer noopener nofollow">[email protected]</a> manager
如果用户 B 执行查询 - select * from people where role = 'manager' limit 1;，那么他应该得到结果:
people_id people_tele people_email role
2 8989898333 <a href="https://stackoverflow.com/cdn-cgi/l/email-protection" class="__cf_email__" data-cfemail="e99a9d8c9f8ca98e84888085c78a8684" rel="noreferrer noopener nofollow">[email protected]</a> manager
方法 1
因此，我想到添加一个 is_processed 列，以免返回相同的结果。因此，在用户 A 执行查询后，该表将如下所示:
people_id people_tele people_email role is_processed
1 8989898332 <a href="https://stackoverflow.com/cdn-cgi/l/email-protection" class="__cf_email__" data-cfemail="751f1a1d1b351218141c195b161a18" rel="noreferrer noopener nofollow">[email protected]</a> manager 1
2 8989898333 <a href="https://stackoverflow.com/cdn-cgi/l/email-protection" class="__cf_email__" data-cfemail="047770617261446369656d682a676b69" rel="noreferrer noopener nofollow">[email protected]</a> manager 0
3 8989898334 <a href="https://stackoverflow.com/cdn-cgi/l/email-protection" class="__cf_email__" data-cfemail="abcac5cfd9cedcebccc6cac2c785c8c4c6" rel="noreferrer noopener nofollow">[email protected]</a> manager 0
4 8989898335 <a href="https://stackoverflow.com/cdn-cgi/l/email-protection" class="__cf_email__" data-cfemail="0265676d70656742656f636b6e2c616d6f" rel="noreferrer noopener nofollow">[email protected]</a> manager 0
方法 2
另一个想法是创建另一个名为 - query_history 的表，其中我有:
query_id people_id processed_time
1 1 22 Jan 2020, 4pm
2 2 22 Jan 2020, 5pm
问题
我的问题是，当用户 A 和用户 B 同时查询时会发生什么？此时系统会返回相同的people_id，并且会向同一个人调用2个电话。
如何解决并发问题？
最佳答案
您只需在方法 1 中添加随机数即可解决该问题
SELECT * FROM people
WHERE role = 'manager'
AND is_processed = 0
order by random()
limit 1;
引用:https://docs.aws.amazon.com/redshift/latest/dg/r_RANDOM.html
关于sql - Redshift 中选择查询的并发性，我们在Stack Overflow上找到一个类似的问题： https://stackoverflow.com/questions/59826303/