Oracle 视图 DBA_WORKLOAD_REPLAYS 官方解释,作用,如何使用详细说明
本站中文解释
Oracle视图DBA_WORKLOAD_REPLAYS定义了可用的回放活动,用于提供有关工作负载回放的信息。
DBA_WORKLOAD_REPLAYS涵盖回放的活动,回放状态和回放结果。可使用此视图来查看工作负载回放正在运行的操作,以及此操作的详细信息,包括回放的初始化和监控状态、正在使用的参数和硬件资源,以及有关回放事件的信息。
使用此视图的步骤如下:
1. 在视图DBA_WORKLOAD_REPLAYS 中查询回放活动信息,
2. 使用回放活动管理器(RAVM)控制看似回放,
3. 查询SQL语句报告,通过ANALYZE_WORKLOAD_REPORTS_PROP_HIST函数实现,
4. 查询文档中的扩展参数状态,
5. 根据报告结果来更新参数状态或重新调整参数,
6. 重新运行回放,
7. 根据新结果来更新参数。
官方英文解释
DBA_WORKLOAD_REPLAYS
displays all the workload replays that have been performed in the current database.
It also lists replays on which DBMS_WORKLOAD_REPLAY.GET_REPLAY_INFO()
has been called. Each row contains information about one workload replay.
Column | Datatype | NULL | Description |
---|---|---|---|
|
|
|
Internal key for the workload replay |
|
|
|
Name of the workload replay |
|
|
|
ID of the database in which the workload was replayed |
|
|
|
Name of the database in which the workload was replayed |
|
|
|
Version of the database in which the workload was replayed |
|
|
Indicates whether the database in which the workload was replayed was an Oracle RAC database ( |
|
|
|
|
Name of the directory object for the workload replay |
|
|
ID of the capture ( |
|
|
|
|
Current status of the workload replay:
|
|
|
Datetime at which the workload prepare started |
|
|
|
Datetime when the replay began |
|
|
|
Datetime when the replay completed or cancelled; NULL if the replay is still in progress |
|
|
|
Duration of the workload replay (in seconds) |
|
|
|
|
Number of workload replay client processes that were used in this workload replay |
|
|
|
Number of workload replay client processes that have finished replay |
|
|
Name of the filter set used for the replay |
|
|
|
|
Reserved for future use |
|
|
Indicates whether recorded transaction semantics should be maintained ( When synchronization is on, the commit order observed during the original workload capture will be preserved. Every action that is replayed will be executed only after all of its dependent commits have been executed. Dependent commits are commits that were issued before the given action in the original workload capture. See Also: |
|
|
|
|
Connection time scaling factor for captured streams during replay. The value is interpreted as a percentage.The default value of 100 means 100 percent. See Also: |
|
|
|
Think time scaling factor for captured streams during replay. It scales the thinking time elapsed between two successive user calls from the same captured stream. The input is interpreted as a percentage. The default value of 100 means 100 percent. See Also: |
|
|
Indicates whether the think time should be automatically corrected between calls ( A value of A value of See Also: |
|
|
|
|
Before the multiple-capture replay, |
|
|
Total number of user calls replayed |
|
|
|
Accumulated database time (in microseconds) for the replay |
|
|
|
Accumulated network time for the replay (in microseconds) |
|
|
|
Accumulated think time (in microseconds) for the replay |
|
|
|
The total time (in seconds) that the replay has been paused (by calling the PAUSE_REPLAY procedure) |
|
|
|
Total number of replayed top-level PL/SQL calls |
|
|
|
Total number of replayed calls for SQL executed from PL/SQL |
|
|
|
Total amount of database time (in microseconds) from PL/SQL calls |
|
|
|
Reserved for future use |
|
|
|
A workload replay uses either the timing information from the capture files or the commit-based synchronization. With commit-based synchronization, the capture-time commit order is preserved during replay, and sessions normally wait on the session that is to do the next commit; such waits are classified as “WCR: replay clock” waits. A replay deadlock occurs if the session that is to do the next commit is itself blocked by a session that is waiting on “WCR: replay clock.” Replay deadlocks are resolved by allowing the blocker to go ahead and commit out of order. Replay deadlocks are not detected as database deadlocks since “WCR: replay clock” is an idle wait, introduced specifically for DB Replay. See Also: “WCR: replay clock” |
|
|
|
Database ID of the AWR snapshots that correspond to this workload replay. For replays that were performed in the current database, this value is equal to the current database’s DBID. For replays that were performed in other databases, this value will either be NULL or will be populated by See Also: DBMS_WORKLOAD_REPLAY.IMPORT_AWR() in Oracle Database PL/SQL |
|
|
|
Begin snapshot ID of the AWR snapshots that correspond to this workload replay |
|
|
|
End snapshot ID of the AWR snapshots that correspond to this workload replay |
|
|
|
Indicates whether the AWR snapshots that correspond to this workload replay have been exported using See Also: DBMS_WORKLOAD_REPLAY.EXPORT_AWR() in Oracle Database PL/SQL |
|
|
|
Error code for this workload replay |
|
|
|
Error message for this workload replay |
|
|
|
|
Full directory path for the replay directory object |
|
|
A hash value computed based on the values of other columns in this view, such as the |
|
|
|
User name of the SQL tuning set owner |
|
|
|
Name of the SQL tuning set for this workload replay |
|
|
|
The name of a schedule to be replayed. It defines one or multiple workload captures, and the order to start their replays. If If See Also: |
|
|
|
Indicates whether replay divergence data have been loaded ( |
|
|
|
Replay options for PL/SQL calls. Possible values:
|
|
|
|
Indicates whether the waiting time for a new session to be connected is automatically reduced when the replay proceeds faster than its capture. The reduced amount is determined by the elapsed-time difference between the replay and the capture of the slowest session. The default value is true. There is no impact when the replay proceeds slower than the capture. |
|
|
|
Replay options in an Oracle RAC environment:
|
|
|
|
Indicates whether only the query-only workload from the current workload capture will be replayed, skipping all the DML/DDL that might update the database ( |
|
|
|
Indicates whether the replay is a PDB-level replay ( |
|
|
|
Number of workload replay clients aborted during the replay |
|
|
|
Database workload capture and replay patch name This column is only populated for backports to Oracle Database 12c Release 1 (12.1) and earlier releases. |
|
|
|
Replay duration limit (in seconds) If the replay duration limit has not been set, then the value of this column is null. |
|
|
|
Provides additional details about the replay status If additional details are not available, then the value of this column is null. |
See Also:
Oracle Database PL/SQL
Packages and Types Reference for more information about the DBMS_WORKLOAD_REPLAY
package