[
https://issues.apache.org/jira/browse/FLINK-24153?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17409499#comment-17409499
]
Ingo Bürk edited comment on FLINK-24153 at 9/3/21, 1:18 PM:
------------------------------------------------------------
In Flink 1.13.1 the SET and RESET commands cannot be used with quotes, this was
only introduced in Flink 1.13.2. You need to use
{code:java}
SET execution.runtime-mode = streaming;
{code}
[~twalthr] Another duplicate of this issue, we should revert the docs changes
for 1.13
was (Author: airblader):
In Flink 1.13.1 the SET and RESET commands cannot be used with quotes, this was
changed in Flink 1.13.2. You need to use
{code:java}
SET execution.runtime-mode = streaming;
{code}
[~twalthr] Another duplicate of this issue, we should revert the docs changes
for 1.13
> Unable to use SET command in Flink SQL Shell file submit mode
> -------------------------------------------------------------
>
> Key: FLINK-24153
> URL: https://issues.apache.org/jira/browse/FLINK-24153
> Project: Flink
> Issue Type: Bug
> Components: Table SQL / Client
> Affects Versions: 1.13.1
> Reporter: Nilanjan Sarkar
> Priority: Major
>
> Let's say you created a Flink SQL job in a file called *sample.sql*
>
> {code:java}
> -- Properties that change the fundamental execution behavior of a table
> program.
> SET 'execution.runtime-mode' = 'streaming';
> SET 'sql-client.execution.result-mode' = 'tableau';
> SET 'sql-client.execution.max-table-result.rows' = '10000';
> SET 'parallelism.default' = '1';
> -- Source table
> CREATE TEMPORARY TABLE salesitems (
> seller_id VARCHAR,
> product VARCHAR,
> quantity INT,
> product_price DOUBLE,
> sale_ts BIGINT,
> proctime AS PROCTIME()
> ) WITH (
> 'connector' = 'kafka',
> 'topic' = 'salesitems',
> 'properties.bootstrap.servers' = '10.129.35.58:29092',
> 'properties.group.id' = 'sliding-windows',
> 'scan.startup.mode' = 'latest-offset',
> 'format' = 'json'
> );
> SELECT
> seller_id,
> HOP_START(proctime, INTERVAL '10' SECONDS, INTERVAL '30' SECONDS) AS
> window_start,
> HOP_END(proctime, INTERVAL '10' SECONDS, INTERVAL '30' SECONDS) AS
> window_end,
> SUM(quantity * product_price) AS window_sales
> FROM salesitems
> GROUP BY
> HOP(proctime, INTERVAL '10' SECONDS, INTERVAL '30' SECONDS),
> seller_id;
> {code}
> Now, if you submit the job as -
> {code:java}
> ${FLINK_HOME}/bin/sql-client.sh embedded \
> --jar ${FLINK_HOME}/lib/flink-sql-connector-kafka_2.11-1.13.1.jar \
> -f sample.sql{code}
> The error you get is -
> {code:java}
> [INFO] Executing SQL from file.
> Flink SQL> -- Properties that change the fundamental execution behavior of a
> table program.
> SET 'execution.runtime-mode' = 'streaming';
> [ERROR] Could not execute SQL statement. Reason:
> org.apache.flink.sql.parser.impl.ParseException: Encountered
> "\'execution.runtime-mode\'" at line 2, column 5.
> Was expecting one of:
> <BRACKET_QUOTED_IDENTIFIER> ...
> <QUOTED_IDENTIFIER> ...
> <BACK_QUOTED_IDENTIFIER> ...
> <HYPHENATED_IDENTIFIER> ...
> <IDENTIFIER> ...
> <UNICODE_QUOTED_IDENTIFIER> ...
> {code}
>
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
