If you use SQL Server 2000, you've likely run into this problem: You want to select a random sampling of rows from a large table with lots of rows, but you're unsure of how to do so. Having a random sampling of rows can be useful when you want to make a smaller version of the table or if you want to troubleshoot a problem by seeing what kinds of rows are in the table.
To get a random sampling, you might be tempted to select the top n rows from the table. However, this sample isn't random and the first n rows aren't necessarily representative of the whole table. Other solutions exist that involve adding columns to the tables, but adding columns isn't always possible or practical.
The standard way to grab random rows from a small table is to use a query such as
ORDER BY NEWID()
The key here is the NEWID function, which generates a globally unique identifier (GUID) in memory for each row. By definition, the GUID is unique and fairly random, so when you sort by that GUID with the ORDER BY clause, you get a random ordering of the rows in the table. Taking the top 10 percent (or whatever percentage you want) will give you a random sampling of the rows in the table.
The NEWID query is often proposed when questions about how to select random rows are asked in discussion groups. It is simple and works very well for small tables. However, the NEWID query has a big drawback when you use it for large tables. The ORDER BY clause causes all the rows in the table to be copied into the tempdb database, where they're sorted. This causes two problems:
- The sorting operation usually has a high cost associated with it. Sorting can use a lot of disk I/O and can run for a long time.
- In the worst-case scenario, tempdb can run out of space. In the best-case scenario, tempdb can take up a large amount of disk space that will never be reclaimed without a manual shrink command.
What you need is a way to randomly select rows that won't use tempdb and won't get much slower as the table gets larger. Here's a new idea on how to do that:
RAND()) as int)) % 100) < 10
The basic idea behind this query is to generate a random number between 0 and 99 for each row in the table, then choose all those rows whose random number is less than the value of the specified percent. In this example, we want about 10 percent of the rows selected randomly, so we choose all rows whose random number is less than 10.
Let's take a closer look at how the (ABS(CAST((BINARY_CHECKSUM(*) * RAND()) as int)) portion of this query works. The BINARY_CHECKSUM function generates a checksum value based on the values of the columns you specify. If two rows are different, they'll typically generate different checksum numbers. The BINARY_CHECKSUM function is generally used to verify whether any of the columns in a row in a table have changed. However, for our purposes, it generates a number that looks like a random number for each row.
The shortcoming of using the BINARY_ CHECKSUM function for our purpose is that every time it's used on a row that hasn't been modified, it returns the same checksum number. Thus, when used by itself, subsequent runs of the query return the same "random" set of rows, which obviously isn't desirable.
To fix this shortcoming, we added the RAND function to the BINARY_ CHECKSUM query. The RAND function scrambles the numbers returned by the BINARY_CHECKSUM function. Thus, we get a different set of rows each time the query is run, making it truly random. The ABS and CAST functions are used because BINARY_CHECKSUM(*) * RAND returns a float that can be a negative number.
The asterisk (*) in BINARY_ CHECKSUM(*) tells the function to use all the columns in the row in its calculations. Alternatively, you can specify a subset of the columns in place of the asterisk. Because this function is CPU-intensive, specifying the minimum number of columns or minimum number of bytes will give you the best performance. The best candidates would be the columns in a unique index. If you decide to use specific columns instead of all the columns, you can add NEWID as a column in the BINARY_CHECKSUM function so that the BINARY_CHECKSUM query will return a random number each time. Thus, you don't need to use RAND in the query, which simplifies it slightly:
(keycol1, NEWID())) as int))
% 100) < 10
Because there's no sorting involved in the BINARY_CHECKSUM query, only a single pass through the table is required to choose n% of the rows. The time and I/O stay linear in proportion to the size of the table.
To test the BINARY_CHECKSUM query against the NEWID query, we set up three large tables containing 1 million rows (435MB), 7 million rows (3GB), and 14 million rows (5.4GB) on a HP ProLiant DL580 G2 server with 1GB memory, four 2.2MHz Intel processors, and eight 36GB disks in RAID 1+0 configuration. Table 1 shows the results. As Table 1 shows, the BINARY_CHECKSUM query saves a lot of time and I/O compared with the NEWID query.
The Microsoft SQL Server team realized that not being able to easily take random samples of rows was a common problem in SQL Server 2000, so the team addressed the problem in SQL Server 2005 by introducing the TABLESAMPLE clause.
This clause selects a subset of rows by choosing random data pages and returning
all the rows on those pages. However, for those of us who still have products
running on SQL Server 2000, who need backward compatibility, or who need truly
row-level randomness, the BINARY_CHECKSUM query is a very effective option.
—Marcelo De Barros and Kenton Gidewall