Confirm the engine is Db2 and read the context before choosing a technique. The need for FROM SYSIBM.SYSDUMMY1 on single-row selects is itself a fingerprint, as is the SYSCAT catalog.

Read the service level and identity. The version and fixpack come from an administrative view, not a function:

' UNION SELECT SERVICE_LEVEL,FIXPACK_NUM FROM SYSIBMADM.ENV_INST_INFO-- 
' UNION SELECT CURRENT USER,CURRENT SERVER FROM SYSIBM.SYSDUMMY1-- 

SYSIBMADM.ENV_SYS_INFO adds host and OS details. The special registers give identity: CURRENT USER (the authorization ID for checks), SESSION_USER and SYSTEM_USER (the session and connected IDs), CURRENT SCHEMA, and CURRENT SERVER (the database).

Map the schema through SYSCAT. List tables, then columns:

' UNION SELECT TABNAME,TABSCHEMA FROM SYSCAT.TABLES-- 
' UNION SELECT COLNAME,TYPENAME FROM SYSCAT.COLUMNS WHERE TABNAME='USERS'-- 

Catalog names are stored upper-case, so match them in upper case. The older SYSIBM.SYSTABLES and SYSIBM.SYSCOLUMNS are alternatives when SYSCAT is filtered. Grants for the current authorization ID come from SYSCAT.DBAUTH and SYSCAT.TABAUTH, which the privileges page uses to decide what the rest of the attack can reach. With the service level, identity, and schema known, extraction proceeds with union, blind, or time-based techniques.

References#

• IBM Db2 SQL Reference: special registers, SYSIBMADM administrative views, SYSCAT catalog
• OWASP Testing Guide: Testing for SQL Injection