Oracle 视图 DBA_WORKLOAD_REPLAYS 官方解释，作用，如何使用详细说明

ID

NUMBER

NOT NULL

Internal key for the workload replay

NAME

VARCHAR2(128)

NOT NULL

Name of the workload replay

DBID

NUMBER

NOT NULL

ID of the database in which the workload was replayed

DBNAME

VARCHAR2(128)

NOT NULL

Name of the database in which the workload was replayed

DBVERSION

VARCHAR2(17)

NOT NULL

Version of the database in which the workload was replayed

PARALLEL

VARCHAR2(3)

Indicates whether the database in which the workload was replayed was an Oracle RAC database (YES) or a single instance database (NO)

DIRECTORY

VARCHAR2(128)

NOT NULL

Name of the directory object for the workload replay

CAPTURE_ID

NUMBER

ID of the capture (DBA_WORKLOAD_CAPTURES.ID) that was replayed. If the replay involves a replay schedule, the CAPTURE_ID will be null.

STATUS

VARCHAR2(40)

NOT NULL

Current status of the workload replay:

• PREPARE – Workload prepare has been started and is waiting for clients to join

• IN PROGRESS – Workload replay is in progress

• COMPLETED – Workload replay has successfully completed

• CANCELLED – Workload replay or the workload prepare has been cancelled

• FAILED – Workload replay was aborted due to errors encountered. See the COMMENTS column for further information.

PREPARE_TIME

DATE

Datetime at which the workload prepare started

START_TIME

DATE

Datetime when the replay began

END_TIME

DATE

Datetime when the replay completed or cancelled; NULL if the replay is still in progress

DURATION_SECS

NUMBER

Duration of the workload replay (in seconds)

NUM_CLIENTS

NUMBER

NOT NULL

Number of workload replay client processes that were used in this workload replay

NUM_CLIENTS_DONE

NUMBER

NOT NULL

Number of workload replay client processes that have finished replay

FILTER_SET_NAME

VARCHAR2(128)

Name of the filter set used for the replay

DEFAULT_ACTION

VARCHAR2(30)

NOT NULL

Reserved for future use

SYNCHRONIZATION

VARCHAR2(9)

Indicates whether recorded transaction semantics should be maintained (TRUE) or not (FALSE)

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: DBMS_WORKLOAD_REPLAY.PREPARE_REPLAY() in Oracle Database PL/SQL
Packages and Types Reference for a detailed explanation of this replay parameter

CONNECT_TIME_SCALE

NUMBER

NOT NULL

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: DBMS_WORKLOAD_REPLAY.PREPARE_REPLAY() in Oracle Database PL/SQL
Packages and Types Reference for a detailed explanation of this replay parameter

THINK_TIME_SCALE

NUMBER

NOT NULL

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: DBMS_WORKLOAD_REPLAY.PREPARE_REPLAY() in Oracle Database PL/SQL
Packages and Types Reference for a detailed explanation of this replay parameter

THINK_TIME_AUTO_CORRECT

VARCHAR2(5)

Indicates whether the think time should be automatically corrected between calls (TRUE) or not (FALSE)

A value of TRUE reduces think time if replay goes slower than capture.

A value of FALSE results in no action.

See Also: DBMS_WORKLOAD_REPLAY.PREPARE_REPLAY() in Oracle Database PL/SQL
Packages and Types Reference for a detailed explanation of this replay parameter

SCALE_UP_MULTIPLIER

NUMBER

NOT NULL

Before the multiple-capture replay, SCALE_UP_MULTIPLIER is used to scale up the query part of a workload capture. The queries from each captured session are replayed concurrently as many times as the value of SCALE_UP_MULTIPLIER.

USER_CALLS

NUMBER

Total number of user calls replayed

DBTIME

NUMBER

Accumulated database time (in microseconds) for the replay

NETWORK_TIME

NUMBER

Accumulated network time for the replay (in microseconds)

THINK_TIME

NUMBER

Accumulated think time (in microseconds) for the replay

PAUSE_TIME

NUMBER

The total time (in seconds) that the replay has been paused (by calling the PAUSE_REPLAY procedure)

PLSQL_CALLS

NUMBER

Total number of replayed top-level PL/SQL calls

PLSQL_SUBCALLS

NUMBER

Total number of replayed calls for SQL executed from PL/SQL

PLSQL_DBTIME

NUMBER

Total amount of database time (in microseconds) from PL/SQL calls

ELAPSED_TIME_DIFF

NUMBER

Reserved for future use

REPLAY_DEADLOCKS

NUMBER

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”

AWR_DBID

NUMBER

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 DBMS_WORKLOAD_REPLAY.IMPORT_AWR().

See Also: DBMS_WORKLOAD_REPLAY.IMPORT_AWR() in Oracle Database PL/SQL
Packages and Types Reference

AWR_BEGIN_SNAP

NUMBER

Begin snapshot ID of the AWR snapshots that correspond to this workload replay

AWR_END_SNAP

NUMBER

End snapshot ID of the AWR snapshots that correspond to this workload replay

AWR_EXPORTED

VARCHAR2(12)

Indicates whether the AWR snapshots that correspond to this workload replay have been exported using DBMS_WORKLOAD_REPLAY.EXPORT_AWR() (YES) or not (NO), or whether AWR snapshots cannot be exported because the replay is still in progress, has run to completion successfully, or was done in a different database from which it was not exported (NOT POSSIBLE)

See Also: DBMS_WORKLOAD_REPLAY.EXPORT_AWR() in Oracle Database PL/SQL
Packages and Types Reference

ERROR_CODE

NUMBER

Error code for this workload replay

ERROR_MESSAGE

VARCHAR2(512)

Error message for this workload replay

DIR_PATH

VARCHAR2(4000)

NOT NULL

Full directory path for the replay directory object

REPLAY_DIR_NUMBER

NUMBER

A hash value computed based on the values of other columns in this view, such as the NAME, DBID, DBNAME, PREPARE_TIME, START_TIME, and END_TIME columns. It should be fairly unique for any replay. The value is used to create a subdirectory under the replay directory.

SQLSET_OWNER

VARCHAR2(128)

User name of the SQL tuning set owner

SQLSET_NAME

VARCHAR2(128)

Name of the SQL tuning set for this workload replay

SCHEDULE_NAME

VARCHAR2(128)

The name of a schedule to be replayed. It defines one or multiple workload captures, and the order to start their replays.

If SCHEDULE_NAME is NULL, the replay is a regular replay introduced since Oracle Database 11g, done with existing APIs from DBMS_WORKLOAD_REPLAY: INITIALIZE_REPLAY, PREPARE_REPLAY, and START_REPLAY.

If SCHEDULE_NAME is not NULL, the replay is a new consolidated replay introduced in Oracle Database 12c. That is, it is a replay of one or more workload captures done with new APIs at DBMS_WORKLOAD_REPLAY: INITIALIZE_CONSOLIDATED_REPLAY, PREPARE_CONSOLIDATED_REPLAY, and START_CONSOLIDATED_REPLAY.

See Also: DBMS_WORKLOAD_REPLAY in Oracle Database PL/SQL
Packages and Types Reference

DIVERGENCE_LOAD_STATUS

VARCHAR2(5)

Indicates whether replay divergence data have been loaded (TRUE) or not (FALSE)

PLSQL_MODE

VARCHAR2(12)

Replay options for PL/SQL calls. Possible values:

• TOP_LEVEL: Top-level PL/SQL only

• EXTENDED: SQL executed from PL/SQL or top-level PL/SQL if there is no SQL recorded inside the PL/SQL

CONNECT_TIME_AUTO_CORRECT

VARCHAR2(12)

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.

RAC_MODE

VARCHAR2(19)

Replay options in an Oracle RAC environment:

• GLOBAL_SYNC: Synchronization is across all instances. This is the default.

  Database connections from workload replay client (wrc) are done based on connection remapping.

• PER_INSTANCE_CLIENT: Synchronization is across all instances.

  All database connections from one wrc are connected to only one instance.

• PER_INSTANCE_SYNC: Synchronization is within one instance.

  All database connections from one wrc are connected to only one instance.

QUERY_ONLY

VARCHAR2(1)

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 (Y) or not (N)

PDB_LEVEL

VARCHAR2(1)

Indicates whether the replay is a PDB-level replay (Y) or not (N)

NUM_CLIENTS_ABORTED

NUMBER

Number of workload replay clients aborted during the replay

PATCH_NAME

VARCHAR2(128)

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.

DURATION_LIMIT

NUMBER(38)

Replay duration limit (in seconds)

If the replay duration limit has not been set, then the value of this column is null.

STATUS_DETAILS

VARCHAR2(128)

Provides additional details about the replay status

If additional details are not available, then the value of this column is null.