· Jul 11

How does SQL Index works in Cache?

Could any one tell me, how the index  works in cache db, Consider for example, I have a table called "Employee" and the fields as "EMPID, EMPNAME, EMPAGE" and I am having index for EMPID as IDX_EMPID and trying to get the record using the following query as


and the table data be like


001       ABC            20

002       AAA            21

003       ABB            23

004       BBB           20

005       BDF           24

006        EEE          22

Will SQL Engine directly hit the "005" record using ^EMPLOYEEI global, if so how it is happening, are they using B-Tree structure?

Please explain

Thanks in Advance

Discussion (2)1
Log in or sign up to continue

As Robert said, the data and indexes are both stored in global trees. The structure of the trees for indices and the trees for general data storage are different, though, and that's where the efficiency comes from.

The data will be stored in the ^EMPLOYEED global. You'll see a bunch of data in that global like:

^EMPLOYEED(1) = $lb("001","ABC",20)

^EMPLOYEED(2) = $ib("002","AAA",21)

Those are lists of every property of that class, indexed by the ID column.

The indexes, on the other hand, will look more like:



This allows for a much more efficient search to find the ID of the records that match your search criteria, similar to if you opened a terminal and put in:

write $ORDER(^EMPLOYEEI("IDX_EMPID","001",""))

Which would give you row 1, then it could more easily look up row 1 in the original without having to go through every record in the table in order searching for it.

In your system management portal, you can run your query and check the query plan, then run it again telling it to ignore indexes like this:


This will tell it to ignore all indexes on the table. You can check that query plan and compare it to the original to see the differences in how the query finds data and how well it performs.