Just adding user@flink.apache.org On Fri, 21 Mar 2025 at 19:08, Kirill Lyashko <kirill.v.lyas...@gmail.com> wrote:
> Ok, I'll ask them as well, thanks. > > On Fri, 21 Mar 2025 at 18:16, Alexey Novakov <ale...@ververica.com> wrote: > >> This is really weird. >> >> I would recommend you also to ask on Flink Dev List. This can be an >> outdated Flink documentation, which shows that multiple statements were >> supported in the past or so. >> >> On Fri, Mar 21, 2025 at 4:35 PM Kirill Lyashko < >> kirill.v.lyas...@gmail.com> wrote: >> >>> This documentation shows such example: >>> >>> https://nightlies.apache.org/flink/flink-docs-release-1.19/docs/dev/table/sql/delete/#delete-statements >>> >>> No, Statement Set not solving the issue as well, I have this problem: >>> >>> Caused by: org.apache.flink.table.api.TableException: Unsupported SQL >>> query! Only accept a single SQL statement of type DELETE. >>> at >>> org.apache.flink.table.api.internal.TableEnvironmentImpl.executeInternal(TableEnvironmentImpl.java:856) >>> ~[flink-table-api-java-uber-1.19.1.jar:1.19.1] >>> at >>> org.apache.flink.table.api.internal.StatementSetImpl.execute(StatementSetImpl.java:109) >>> ~[flink-table-api-java-uber-1.19.1.jar:1.19.1] >>> >>> What is similar to multiple queries joined with ';' in one string. >>> >>> Ok, looks like it's not supported, thanks for help. >>> >>> On Fri, 21 Mar 2025 at 12:33, Alexey Novakov <ale...@ververica.com> >>> wrote: >>> >>>> I think documentation talks more about SQL client / SQL gateway where >>>> we can run DML statements one by one. >>>> I have not found a clear example in Table API documentation with more >>>> than one query execution within the same Flink application. >>>> >>>> Not sure about DELETEs, need to test. >>>> >>>> Best regards, >>>> Alexey >>>> >>>> >>>> On Fri, Mar 21, 2025 at 11:35 AM Kirill Lyashko < >>>> kirill.v.lyas...@gmail.com> wrote: >>>> >>>>> Ok, but in that case the documentation is saying the opposite, no? >>>>> >>>>> Thanks, for recommendation, I will take a look, but from the first >>>>> glance it looks like it supports only INSERTS, or not? >>>>> >>>>> On Fri, 21 Mar 2025 at 11:23, Alexey Novakov <ale...@ververica.com> >>>>> wrote: >>>>> >>>>>> Hi Kirill, >>>>>> >>>>>> As far as I know, multiple statements cannot be run by calling an >>>>>> execute() method on each, so you are missing nothing. >>>>>> Try to look at the Statement Set which allows you to submit multiple >>>>>> SQL statements within one Flink app. >>>>>> >>>>>> For example: >>>>>> https://github.com/novakov-alexey/flink-ml-sandbox/blob/main/src/main/scala/com/example/customerChurn.scala#L223 >>>>>> >>>>>> Best regards, >>>>>> Alexey >>>>>> >>>>>> On Thu, Mar 20, 2025 at 4:57 PM Kirill Lyashko < >>>>>> kirill.v.lyas...@gmail.com> wrote: >>>>>> >>>>>>> I' m trying to submit multiple DELETE statements in to the >>>>>>> TableEnvironment in way: >>>>>>> >>>>>>> val settings = EnvironmentSettings.newInstance.inBatchMode.build() >>>>>>> val env = TableEnvironment.create(settings) >>>>>>> createSchema(env) >>>>>>> val queries = getDeleteQueries >>>>>>> queries.foreach(q => env.executeSql(q).await()) >>>>>>> >>>>>>> what according to documentation should just work delete-statements >>>>>>> <https://nightlies.apache.org/flink/flink-docs-release-1.19/docs/dev/table/sql/delete/#delete-statements>. >>>>>>> But in reality I have follow exception after the first query is >>>>>>> finished: >>>>>>> >>>>>>> Caused by: org.apache.flink.util.FlinkRuntimeException: Cannot have >>>>>>> more than one execute() or executeAsync() call in a single environment. >>>>>>> at >>>>>>> org.apache.flink.client.program.StreamContextEnvironment.validateAllowedExecution(StreamContextEnvironment.java:199) >>>>>>> ~[flink-dist-1.19.1.jar:1.19.1] >>>>>>> at >>>>>>> org.apache.flink.client.program.StreamContextEnvironment.executeAsync(StreamContextEnvironment.java:187) >>>>>>> ~[flink-dist-1.19.1.jar:1.19.1] >>>>>>> at >>>>>>> org.apache.flink.table.planner.delegation.DefaultExecutor.executeAsync(DefaultExecutor.java:110) >>>>>>> ~[?:?] >>>>>>> at >>>>>>> org.apache.flink.table.api.internal.TableEnvironmentImpl.executeInternal(TableEnvironmentImpl.java:1032) >>>>>>> ~[flink-table-api-java-uber-1.19.1.jar:1.19.1] >>>>>>> at >>>>>>> org.apache.flink.table.api.internal.TableEnvironmentImpl.executeInternal(TableEnvironmentImpl.java:876) >>>>>>> ~[flink-table-api-java-uber-1.19.1.jar:1.19.1] >>>>>>> at >>>>>>> org.apache.flink.table.api.internal.TableEnvironmentImpl.executeInternal(TableEnvironmentImpl.java:1112) >>>>>>> ~[flink-table-api-java-uber-1.19.1.jar:1.19.1] >>>>>>> at >>>>>>> org.apache.flink.table.api.internal.TableEnvironmentImpl.executeSql(TableEnvironmentImpl.java:735) >>>>>>> ~[flink-table-api-java-uber-1.19.1.jar:1.19.1] >>>>>>> >>>>>>> When I check the code I see that there is indeed a validation that >>>>>>> not more then 1 job can be submitted: >>>>>>> >>>>>>> private void validateAllowedExecution() { >>>>>>> if (enforceSingleJobExecution && jobCounter > 0) { >>>>>>> throw new FlinkRuntimeException( >>>>>>> "Cannot have more than one execute() or executeAsync() >>>>>>> call in a single environment."); >>>>>>> } >>>>>>> jobCounter++; >>>>>>> } >>>>>>> >>>>>>> When I've tried to submit multiple statements separated by ; I've >>>>>>> got follow exception: >>>>>>> >>>>>>> Caused by: java.lang.IllegalArgumentException: only single statement >>>>>>> supported >>>>>>> at >>>>>>> org.apache.flink.util.Preconditions.checkArgument(Preconditions.java:138) >>>>>>> ~[flink-dist-1.19.1.jar:1.19.1] >>>>>>> at >>>>>>> org.apache.flink.table.planner.delegation.ParserImpl.parse(ParserImpl.java:104) >>>>>>> ~[?:?] >>>>>>> at >>>>>>> org.apache.flink.table.api.internal.TableEnvironmentImpl.executeSql(TableEnvironmentImpl.java:728) >>>>>>> ~[flink-table-api-java-uber-1.19.1.jar:1.19.1] >>>>>>> >>>>>>> My job is submitted to Kubernetes vie usage of >>>>>>> flink-kubernetes-operator >>>>>>> <https://nightlies.apache.org/flink/flink-kubernetes-operator-docs-release-1.10/> >>>>>>> >>>>>>> Flink version is 1.19.1 >>>>>>> >>>>>>> Flink Kubernetes operator version is 1.10.0 >>>>>>> >>>>>>> My queries look like: >>>>>>> >>>>>>> DELETE FROM hdfs_table >>>>>>> WHERE id IN ( >>>>>>> SELECT DISTINCT a.id >>>>>>> FROM jdbc_table_a a >>>>>>> JOIN jdbc_table_b b ON a.b_id = b.id >>>>>>> WHERE b.should_be_deleted >>>>>>> ) >>>>>>> >>>>>>> Two questions which I would appreciate to get help with: >>>>>>> >>>>>>> 1. What am I missing in the configuration that's preventing me >>>>>>> from submitting multiple DELETE statements? >>>>>>> 2. Is it possible to submit multiple DELETE questions in >>>>>>> parallel to be able to reuse jdbc tables rather than load them on >>>>>>> every >>>>>>> query? >>>>>>> >>>>>>>
