Skip to content

[Bug] S3 load fails when access GCS #20589

@michael1991

Description

@michael1991

Search before asking

  • I had searched in the issues and found no similar issues.

Version

1.2.4.1

What's Wrong?

When I use S3 load or S3 tvf to load parquet files stored in GCS, I get errors as below:

2023-06-08 07:14:51,734 ERROR (mysql-nio-pool-3539|101239) [S3Storage.list():372] errors while get file status
org.apache.doris.common.UserException: errCode = 2, detailMessage = Failed to get S3 FileSystem for bucket is null/empty
	at org.apache.doris.backup.S3Storage.getFileSystem(S3Storage.java:142) ~[doris-fe.jar:1.2-SNAPSHOT]
	at org.apache.doris.backup.S3Storage.list(S3Storage.java:356) ~[doris-fe.jar:1.2-SNAPSHOT]
	at org.apache.doris.common.util.BrokerUtil.parseFile(BrokerUtil.java:90) ~[doris-fe.jar:1.2-SNAPSHOT]
	at org.apache.doris.tablefunction.ExternalFileTableValuedFunction.parseFile(ExternalFileTableValuedFunction.java:142) ~[doris-fe.jar:1.2-SNAPSHOT]
	at org.apache.doris.tablefunction.S3TableValuedFunction.<init>(S3TableValuedFunction.java:127) ~[doris-fe.jar:1.2-SNAPSHOT]
	at org.apache.doris.tablefunction.TableValuedFunctionIf.getTableFunction(TableValuedFunctionIf.java:49) ~[doris-fe.jar:1.2-SNAPSHOT]
	at org.apache.doris.analysis.TableValuedFunctionRef.<init>(TableValuedFunctionRef.java:40) ~[doris-fe.jar:1.2-SNAPSHOT]
	at org.apache.doris.analysis.CUP$SqlParser$actions.case817(SqlParser.java:28107) ~[doris-fe.jar:1.2-SNAPSHOT]
	at org.apache.doris.analysis.CUP$SqlParser$actions.CUP$SqlParser$do_action(SqlParser.java:8620) ~[doris-fe.jar:1.2-SNAPSHOT]
	at org.apache.doris.analysis.SqlParser.do_action(SqlParser.java:2291) ~[doris-fe.jar:1.2-SNAPSHOT]
	at java_cup.runtime.lr_parser.parse(lr_parser.java:584) ~[jflex-1.4.3.jar:?]
	at org.apache.doris.common.util.SqlParserUtils.getMultiStmts(SqlParserUtils.java:60) ~[doris-fe.jar:1.2-SNAPSHOT]
	at org.apache.doris.qe.ConnectProcessor.parse(ConnectProcessor.java:388) ~[doris-fe.jar:1.2-SNAPSHOT]
	at org.apache.doris.qe.ConnectProcessor.handleQuery(ConnectProcessor.java:285) ~[doris-fe.jar:1.2-SNAPSHOT]
	at org.apache.doris.qe.ConnectProcessor.dispatch(ConnectProcessor.java:473) ~[doris-fe.jar:1.2-SNAPSHOT]
	at org.apache.doris.qe.ConnectProcessor.processOnce(ConnectProcessor.java:700) ~[doris-fe.jar:1.2-SNAPSHOT]
	at org.apache.doris.mysql.ReadListener.lambda$handleEvent$0(ReadListener.java:52) ~[doris-fe.jar:1.2-SNAPSHOT]
	at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) ~[?:1.8.0_372]
	at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) ~[?:1.8.0_372]
	at java.lang.Thread.run(Thread.java:750) ~[?:1.8.0_372]
Caused by: java.lang.IllegalArgumentException: bucket is null/empty
	at org.apache.hadoop.thirdparty.com.google.common.base.Preconditions.checkArgument(Preconditions.java:144) ~[hadoop-shaded-guava-1.1.1.jar:1.1.1]
	at org.apache.hadoop.fs.s3a.S3AUtils.propagateBucketOptions(S3AUtils.java:1162) ~[hadoop-aws-3.3.3.jar:?]
	at org.apache.hadoop.fs.s3a.S3AFileSystem.initialize(S3AFileSystem.java:426) ~[hadoop-aws-3.3.3.jar:?]
	at org.apache.hadoop.fs.FileSystem.createFileSystem(FileSystem.java:3469) ~[hadoop-common-3.3.3.jar:?]
	at org.apache.hadoop.fs.FileSystem.get(FileSystem.java:537) ~[hadoop-common-3.3.3.jar:?]
	at org.apache.doris.backup.S3Storage.getFileSystem(S3Storage.java:140) ~[doris-fe.jar:1.2-SNAPSHOT]
	... 19 more

Load SQL

LOAD LABEL label_gcs_test 
( 
    DATA INFILE("s3://bucket/tables_prod/parquet_log_report/part*") 
    INTO TABLE load_test 
    FORMAT AS "parquet"
) 
WITH S3
( 
    "AWS_ENDPOINT" = "https://storage.googleapis.com", 
    "AWS_ACCESS_KEY" = "ak", 
    "AWS_SECRET_KEY" = "sk", 
    "AWS_REGION" = "us"
) 
PROPERTIES ( "timeout" = "3600" );

LOAD LABEL label_gcs_test 
( 
    DATA INFILE("https://storage.googleapis.com/bucket/tables_prod/parquet_log_report/part*") 
    INTO TABLE load_test 
    FORMAT AS "parquet"
) 
WITH S3
( 
    "AWS_ENDPOINT" = "https://storage.googleapis.com", 
    "AWS_ACCESS_KEY" = "ak", 
    "AWS_SECRET_KEY" = "sk", 
    "AWS_REGION" = "us",
    "use_path_style" = "true"
) 
PROPERTIES ( "timeout" = "3600" );

Whatever which method I used, I always got same error. Then I tried in code:

URI uri = new URI("s3://bucket/tables_prod/parquet_log_report/part*");

Then bucket ( uri.getHost() ) is null.

What You Expected?

S3 load can load data stored in GCS successfully.

How to Reproduce?

No response

Anything Else?

No response

Are you willing to submit PR?

  • Yes I am willing to submit a PR!

Code of Conduct

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions