You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This will be happen only on single routing which do not rewrite the logical SQL's Limit sub clause.
5 rows is skipped. 5 is the offset of Limit sub clause. This is duo to offset skipping procedure in merge step.
When single routing, offset should not be skipped.
Steps to reproduce the behavior, such as: SQL to execute, sharding rule configuration, when exception occur etc.
Take ShardingType.SHARDING_DATABASES for example:
The SQL is single routed to demo_ds_0.t_order.
SELECT * FROM t_order WHERE user_id = 0 LIMIT 5, 10;
Example codes for reproduce this issue (such as a github link).
Offical example.
The text was updated successfully, but these errors were encountered:
Bug Report
Which version of ShardingSphere did you use?
3.0.1-SNAPSHOT
Which project did you use? Sharding-JDBC or Sharding-Proxy?
Sharding-JDBC and Sharding-Proxy
Expected behavior
Take ShardingType.SHARDING_DATABASES for example:
The SQL is single routed to demo_ds_0.t_order.
SELECT * FROM t_order WHERE user_id = 0 LIMIT 5, 10;
10 rows are expected:
order_id: 0, user_id: 0, status: 6
order_id: 0, user_id: 0, status: 7
order_id: 0, user_id: 0, status: 8
order_id: 0, user_id: 0, status: 9
order_id: 0, user_id: 0, status: 10
order_id: 0, user_id: 0, status: 11
order_id: 0, user_id: 0, status: 12
order_id: 0, user_id: 0, status: 13
order_id: 0, user_id: 0, status: 14
order_id: 0, user_id: 0, status: 15
Actual behavior
Get 5 rows result:
order_id: 0, user_id: 0, status: 11
order_id: 0, user_id: 0, status: 12
order_id: 0, user_id: 0, status: 13
order_id: 0, user_id: 0, status: 14
order_id: 0, user_id: 0, status: 15
Reason analyze (If you can)
This will be happen only on single routing which do not rewrite the logical SQL's Limit sub clause.
5 rows is skipped. 5 is the offset of Limit sub clause. This is duo to offset skipping procedure in merge step.
When single routing, offset should not be skipped.
Steps to reproduce the behavior, such as: SQL to execute, sharding rule configuration, when exception occur etc.
Take ShardingType.SHARDING_DATABASES for example:
The SQL is single routed to demo_ds_0.t_order.
SELECT * FROM t_order WHERE user_id = 0 LIMIT 5, 10;
Example codes for reproduce this issue (such as a github link).
Offical example.
The text was updated successfully, but these errors were encountered: