1. How does one calculate data?Data Statistics are used to record the activity of data statistics, including data statistics and SQL statements, log on to Alibaba Cloud dataworks.
Ii. Memory storage method of memory recordThere are two types: memory in the operating system file and SYS. AUD $ in the system table space.
Iii. Test results of performance impact on Data ProcessingCPU usage is inevitable. Therefore, it is necessary to combine the balance between the demand and performance to determine the best tuning strategy.
4. What information does the result contain?There are two methods for storing data and file storage, as shown in the preceding figure. For scenarios where the data volume is stored, sys. the AUD $ table contains the following information: · Operating System using anonymous users using voice recognition-graphic client standards-operations on image names using the compromised Schema · complete replacement of Operations · date and time stamp if the log is stored in an external operation file, the file may contain the following information: · Operation System-generated computing records · digital computing records · calculated data operations · hypervisor (sys) stored in a file is stored as a memory. If you want to understand the information corresponding to these operations, you can query the following data dictionary tables: data dictionary table ------------------------------------- the data dictionary operation Representative indicates the operation of the data dictionary. Its correlation description can be used to check the permission used by the Operation in the audit_actions table of the data dictionary table. The permission description can be used in the system_privilege_map dictionary table to explain the complete operation token. 0 is returned when the operation is successful, when the loss occurs, the relevant Oracle response information is returned. The statement corresponding to the response information can be obtained from the official Oracle documents --------------------------------
5. Default MetricRegardless of whether the computing function has been enabled before, some data-related operations will be stored in the external computing file (note: not into sys. AUD $ table). The default operations are as follows: connect as sysdba or connect as sysoper) · data migration failed · data migration stopped
Vi. billing of operations performed by hypervisor using tokensThe hypervisor uses the as sysdba or as sysoper method to access data centers. The initialization metrics audit_sys_operations are used to specify the metrics for the hypervisor. If you set audit_sys_operations to true, then, all operations of the hypervisor will be scheduled, regardless of whether or not the scaling function is enabled, all metric information is stored in the external metric file. AUD $ table ).
7. histogram Calculation
1. Set the storage location of the log ServiceThe initialization metric data audit_trail specifies the storage location of the metric data. The metric data has three values: DB metric data is stored in the SYS of the metric data. in the AUD $ table, the OS memory is stored in the external operating system file, and the non-None relationship calculation (default) the initialization metric data audit_file_dest specifies the default path for storing the internal external log file. The default value is $ ORACLE_HOME/rdbms/audit.
2. Program SelectionThis option indicates the availability of the audit statement used by the online statement. For the audit method of the audit statement of audit, please refer to the official Oracle SQL audit.
3. Give a clear picture in practiceA. calculate the audit session by Scott, Fey; B. calculate the creation and end of audit sessions for all sessions that use the role; C. it is estimated that the operation for table Division audit delete any table; C. calculate the operation of table Division (Restriction: only allow the calculation of the loss condition) Audit delete any table whenever not successful; D. maxcompute calculates the operation of table Division (Restriction: Only Calculation of successful division conditions) Audit delete any table whenever successful; E. calculate the operation except the table (Limit: the same operation sentence in the same session is only counted once) Audit delete any table by session; (Note: in this case, if the computer is set to be stored in an external file, this effect cannot be realized, because Oracle cannot determine whether the same operation sentence has been computed .) F. operation of table Division calculation (Restriction: All statements for table division are calculated, regardless of whether there are multiple identical operation sentence in the same session) audit delete any table by access; G. calculate Fey. the delete operation of the employee table audit delete on Fey. employee; H. calculate Fey. delete, updet, insert audit Delete, update, insert on Fey. employee;
8. Stop planningStop using the noaudit statement. For the method of writing the statement, refer to the Oracle official SQL statement examination. The following example shows how to use noaudit:. stop the noaudit session for all session creation and end; B. stop creating and ending noaudit session by fey, Scott; C. noaudit delete any table; C. stop the operation of maxcompute table Division (Restriction: Stop the operation of maxcompute table Division) noaudit delete any table whenever not successful; D. stop the operation of maxcompute table Division (Restriction: Stop the operation when the operation is successful) noaudit delete any table whenever successful; G. stop the program to calculate Fey. delete noaudit delete on fey in the employee table. employee; H. stop the program to calculate Fey. delete, updet, insert noaudit Delete, update, insert on Fey. employee;
9. Statement of memory
SYS. AUD $
ManagementWhen sys. an AUD $ table has more and more memory records, and even when the memory limit is reached, the cause is that the website cannot be written. Maybe I still need to sort the data in the table, or sort the data into shards that we think are no longer needed. These are all tests. Below is a method for sorting the shards in the Table:. select into to another table, or use export to export data to external file B. truncate. then, we need to add the original data volume. Maybe we need to calculate the operation on the SYS. AUD $ table, for example, audit insert, update, deleteon SYS. AUD $ by access;
10. fine-grained statisticsAs described above, the log does not contain the number of operations. For example, you need to calculate the SELECT statement for a table, in addition, the data returned by the SELECT statement must be included in the log. In this case, you need to use the fine-grained program. Fine-grained events are Oracle-based events. The principle is to intercept related data records in insert, update, or delete events. It is caused by the norm of the Orale program. You can refer to Oracle documents for more information.
11. computation-related data dictionaryAccording to the statement in each region of mongodba_audit_trail, please refer to the official Oracle expert manual.
