Identifying and Correcting Problems
Use
The initial screen of SM56 displays the buffer and server times in the statistics information. You can see the number of accesses and their duration. You can also identify any possible problems or bottlenecks.
Provided only small numbers are displayed on the right-hand side, everything is working correctly. If there are a lot of entries in the column with >=, you probably have too many server calls. You can solve this by increasing the numbers in the buffer.
Unfortunately you cannot tell from these numbers which intervals are causing the problems. You must then proceed heuristically with the analysis. You can run an SQL trace on the server to only trace accesses to the NRIV table. In the trace you can identify the number range interval for which the database must frequently be accessed. You can then increase the number of numbers in the buffer for the associated number range object in transaction SNRO.