Page tree
Skip to end of metadata
Go to start of metadata

Transaction SM56 ( Number Range Buffer) is used for the administration of the number range buffer, and provides an overview of available information on the number range buffer on each application server.
It's used when Number Range object has defined buffering method as "Main Memory" (SNRO transaction).

The top screen area contains general information about the statistics.

InformationMeaning

Max. No. of Entries

Maximum number of entries in the buffer (see the nobuf/max_no_buffer_entries profile parameter)

Current No. of Entries

Current number of entries in the buffer

Max. No. of Indexes

Size of an internal administration structure. Normally you can see three times as many entries here than in the actual buffer.

Current No. of Indexes

Number of entries in the internal administration structure.

Size

Memory requirement of the buffer in bytes

Buffer Calls

Number of calls made to the buffer. (There are other functions too, which are not counted. These are also used by the NUMBER_GET_NEXT function group).

Get Calls

Number of calls that are counted

Server Calls

Number of calls made to the number range server

Database Calls

There are also database calls that are not made within the context of the number range server. These calls are included in the count here.

Conflicts

SAP internal value for error analysis

Timeouts

If a process wants to assign a number and has to wait because another process is blocking the lock of the buffer or interval, it waits for a specified length of time ( nobuf/max_attempts in seconds). Once this time has elapsed, an error is notified to NUMBER_GET_NEXT. The process then tries to assign a number directly from the database. This value shows how often this timeout has occurred.

Therefore, this field should only provide a SAP internal value for error analysis.
Sometimes the value from this field is different from 0; but this does not indicate any error.

  • When a new entry should be created, this can be done via a linear search through the list or with some sort of index based algorithm, which is usually faster. 
  • So, when searching for a free entry in the number range buffer, normally an algorithm is used to detect a free entry.
  • If the entry to be inserted is too similar to an already existing entry, the algorithm reveals the same result and might return the same index. Therefore, this produces a "Conflict". 
  • In that case, a linear algorithm is used to detect a free entry instead.

So there are no functional consequences nor performance issues. The line has a more statistical character then indicating a problem.

  • No labels