Striim 3.9.4 / 3.9.5 documentation

Console commands

If Striim is installed in /opt, the command to run the console is:

/opt/Striim/bin/ -c <cluster name>

The following switches may be used:

-c <cluster name>: the name of the cluster to connect to (if omitted, will default to the current user name)

-f <path><file name>.tql: a TQL file containing commands to run when console starts (if not specified from root, path is relative to the Striim program directory)

-i <IP address>: the IP address of the Striim server to connect to

-p <password>: the password to use to log in (if omitted, you will be prompted after the console connects)

-u <user name>: the user to log in as (if omitted, you will be prompted after the console connects)

The following commands are intended primarily for use at the command line rather than in .tql application files. DDL and component reference and ad-hoc queries may also be entered at the command line.


@<path>/<file name>.tql;

Run the commands in the specified TQL file (typically all the commands required to create an application). If not specified from root, path is relative to the Striim program directory. Note that if the file contains DDL defining an application, the name will be defined by the CREATE APPLICATION statement, not the TQL file name.


CONNECT <user name> <password> [ TO <cluster name> ];


WACTIONSTORE | WINDOW } <namespace>.<object name>;

Returns all properties of the specified component, application, or flow. DESCRIBE CLUSTER; returns the cluster name, information about the metadata repository, and license details. 


Deploys an application in the Created state. See Managing deployment groups.


See Reading a Kafka stream with an external Kafka consumer.



Lists all commands previously entered in the current console session.



Returns a list of all objects of the specified type.


See Loading standalone sources, caches, and WActionStores.


See Using the MON command.


QUIESCE <application name>;
  1. Pauses all sources.

  2. Flushes out all data in process. This can result in partial batches of events, such as a 100-event window emitting a batch of only 20 events, or a five-minute window emitting a one-minute batch, which can result in results from functions such as COUNT and SUM being outside of the normal range.

  3. After all data is flushed, records all information required for recovery, if it is enabled (see Recovering applications).

  4. Stops the application.

The primary use for QUIESCE is prior to using ALTER and RECOMPILE on an application with recovery enabled. Because the recovery information is recorded when the application contains no data in process, it can be modified in any way other than changing the sources, and when it is restarted it will resume operation with no duplicate events. If the sources are recoverable (see Recovering applications), there will also be no missing events ("exactly-once processing"). The tradeoff for this is the potential for out-of-normal-range result. If instead you use STOP, there will be no anomalous results, but after restarting the application you may have some missing or duplicate events.

Due to long-running open transactions, OracleReader may be unable to pause, in which case the application and all its sources will resume as if the QUIESCE command had not been issued.

To support QUIESCE with OracleReader, see Creating the quiescemarker table.


Seee Using the REPORT START / STOP command.


When an application is in the CRASHED state and the condition that caused it to crash has been corrected, RESUME <application name>; will attempt to resume operation from the point where the application crashed.


Queries a WActionStore. See Browsing data with ad-hoc queries.


SHOW <stream name>;

Returns the output of a stream. Press Ctrl-D to end.

SHOW <source or target name> LINEAGE
 [LIMIT <count>] 
  [-start  'yyyy-mm-dd']
  [-end  'yyyy-mm-dd']

For file-based readers and writers, returns a list of files read or created. For example:


By default, the most recent ten files are listed. You can change that number by adding the LIMIT option to the command: for example, LIMIT 25. To see the oldest files instead of the newest, add the DESC option. Use the -status-start, and/or -end options to show only a specified subset of the files.

For OracleReader see File lineage in Oracle.

By default, this feature is disabled. See Enabling file lineage.

With OracleReader, SHOW can also be used for Viewing open transactions.


START <application name>;

Starts the application.


STOP <application name>;

Stops a running application and, if recovery is enabled, writes recovery checkpoints. Any data currently in the local buffers used by BigQueryWriter, RedshiftWriter, or S3Writer will be written when the application is restarted, which may increase the amount of time it takes to restart. Any other data currently being processed will be lost: to avoid this, use QUIESCE instead. If recovery is enabled and  the application was not dropped after it was stopped, recoverable sources will attempt to restart where they left off (see Recovering applications).

The Stop command is also available when an application is in the Starting state. This can be useful when an application is stuck in the Starting state and you want to stop and debug it rather than waiting for it to time out and revert to the Deployed state.


Undeploys a deployed application. See Managing deployment groups.

The Undeploy command is also available when an application is in the Deploying state. This can be useful when an application is stuck in the Deploying state and you want to stop and debug it rather than waiting for it to time out and revert to the Created state.


See Loading standalone sources, caches, and WActionStores.


USAGE [<application name>];

Lists sources and how much total data each has acquired. If you do not specify an application name, USAGE; lists all sources for all applications.