I have a typical table of hierarchical data in id, parentId form.
CREATE TABLE Hierarchy (Id int, ParentId int NULL, Name varchar(128));
INSERT INTO Hierarchy VALUES (1, NULL, '1');
INSERT INTO Hierarchy VALUES (2, NULL, '2');
INSERT INTO Hierarchy VALUES (3, NULL, '3');
INSERT INTO Hierarchy VALUES (4, 1, '1.1');
INSERT INTO Hierarchy VALUES (5, 1, '1.2');
INSERT INTO Hierarchy VALUES (6, 4, '1.1.1');
I need to detect cycles like below in the existing data.
Id ParentId Name
27 8 'foo'
8 19 'bar'
19 27 'busted'
The Ids are from another table so I cannot use ordering of the Ids as part of the solution. There are about 1/2 million rows. The data is a large number of independent trees of height 1-5. The intent is to clean up the data to eliminate the cycles then add a trigger to the table to prevent cycles in the future.
I know the algorithms for cycle detection. It seems to me like it should be a common problem. So, I was wondering if there was a handy way in the SQL Server `box' to accomplish this with a minimum of code.