Community Note

• Please vote on this issue by adding a 👍 reaction
  to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do
  not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment.

Debug Output

Caused by: liquibase.exception.DatabaseException: Connection could not be created to jdbc:bigquery://https://www.googleapis.com/bigquery/v2:443;ProjectId=<project_id>;Location=EU;OAuthType=3;DefaultDataset=himanshu_local_cdc;Timeout=3600; with driver com.simba.googlebigquery.jdbc.Driver.  [Simba][BigQueryJDBCDriver](100004) HttpTransport IO error : Error getting subject token from metadata server: PKIX path building failed: sun.security.provider.certpath.SunCertPathBuilderException: unable to find valid certification path to requested target.
	at liquibase.ext.bigquery.database.BigqueryConnection.openConn(BigqueryConnection.java:106)
	at liquibase.ext.bigquery.database.BigqueryConnection.open(BigqueryConnection.java:94)
	at liquibase.database.ConnectionServiceFactory.create(ConnectionServiceFactory.java:32)
	... 30 more

Panic Output

NA

Steps to Reproduce

1. Use Liquibase 4.28.0 in github action as container
2. Download bigquery jdbc jars and liquibase-bigquery jar and copy it inside liquibase/lib
3. Use github action google-github-actions/auth@v2 to setup WIF, this service account has bigquery admin permissions
4. Finally run liquibase using command below

liquibase \
            --driver=com.simba.googlebigquery.jdbc.Driver \
            --url="jdbc:bigquery://https://www.googleapis.com/bigquery/v2:443;ProjectId=<project_id>;Location=EU;OAuthType=3;DefaultDataset=himanshu_local_cdc;Timeout=3600;" \
            --changeLogFile="changelog.xml" \
            --searchPath="./liquibase/cdc-tables" \
            --log-level debug \
            update

5. I have also tried using encrypt=false;trustServerCertificate=true; as suggested on post here even thought it was for sqlserver but error was on the same lines but still same issue.

Expected Behavior

Liquibase changeset should have got executed

Actual Behavior

Error below

Caused by: liquibase.exception.DatabaseException: Connection could not be created to jdbc:bigquery://https://www.googleapis.com/bigquery/v2:443;ProjectId=<project_id>Location=EU;OAuthType=3;DefaultDataset=himanshu_local_cdc;Timeout=3600; with driver com.simba.googlebigquery.jdbc.Driver.  [Simba][BigQueryJDBCDriver](100004) HttpTransport IO error : Error getting subject token from metadata server: PKIX path building failed: sun.security.provider.certpath.SunCertPathBuilderException: unable to find valid certification path to requested target.
	at liquibase.ext.bigquery.database.BigqueryConnection.openConn(BigqueryConnection.java:106)
	at liquibase.ext.bigquery.database.BigqueryConnection.open(BigqueryConnection.java:94)
	at liquibase.database.ConnectionServiceFactory.create(ConnectionServiceFactory.java:32)
	... 30 more

References

• #0000

Community Note

Bigquery BigQuerySequenceSnapshotGenerator register itself with a higher priority than other extensions even when we are not using BigQuery database - just having the jar at lib folder does it. It needs the same fix as described on Snowflake PR liquibase/liquibase#3572 (review).

When creating Stored Procedures containing CASE statements like below, Liquibase throws liquibase.exception.DatabaseException and fails with EXEC_JOB_EXECUTION_ERR

Sample Stored Procedure

Note: All the referenced tables, udfs are already in place.

--liquibase formatted sql
--changeset gcp:1 runOnChange: true

CREATE OR REPLACE PROCEDURE `project.dataset.sp`(load_dt DATE)
BEGIN

  BEGIN TRANSACTION;

 --Full Refresh Table truncate
  TRUNCATE TABLE `project.dataset.sp`;

  --Insert into structured table
  INSERT INTO `project.dataset.sp` (
  SELECT
    `project.dataset.sp.udf_std_str_to_str`(epvid_evt_id) as event_id,
    `project.dataset.sp.udf_std_str_to_str`(epvid_vend_sbsy) as vendor_subsidiary,
    CASE WHEN epvid_vers_id LIKE '%-%' THEN NULL ELSE `project.dataset.sp.udf_std_str_to_str`(epvid_vers_id) END as version_id,
    `project.dataset.sp.udf_std_str_to_str`(epvid_itm_nbr) as item_nbr,
    `project.dataset.sp.udf_std_str_to_date`(epvid_trn_dte,'%Y-%m-%d') as transaction_date,
    `project.dataset.sp.udf_std_str_to_str`(epvid_rgn) as region_nbr,
    `project.dataset.sp.udf_std_str_to_str`(epvid_cur_itm_nbr) as current_item_nbr,
    epvid_act_itsl_dlr as actual_sale_dlr,
    epvid_act_itsl_unt as actual_sale_units,
    epvid_act_itmd_dlr as actual_markdown_dlr,
    epvid_act_itgm_dlr as actual_gross_margin_dlr,
    `project.dataset.sp.udf_std_str_to_str`(epvid_crt_opr_id) as create_operator_id,
    `project.dataset.sp.udf_std_str_to_date`(epvid_crt_dte,'%Y-%m-%d') as create_date,
    `project.dataset.sp.udf_std_str_to_datetime`(epvid_lst_mnt_tsmp,'%Y-%m-%d %H:%M:%E*S') as last_maintenance_timestamp,
    `project.dataset.sp.udf_std_str_to_str`(epvid_lst_opr_id) as last_op_id,
    `project.dataset.sp.udf_std_str_to_str`(epvid_lst_trn_cde) as last_trnsctn_code,
    `project.dataset.sp.udf_std_str_to_str`(epvid_grp_id) as grp_id,
    epvid_out_stk_ind as out_of_stock_ind,
    epvid_nbr_str_out as number_of_stores_out,
    `project.dataset.sp.udf_std_str_to_str`(epvid_pvt_itlb_ind) as pvt_lbl_ind,
    file_ingested_date,
    "dag_struct_load_evt_item_daily" as dag_id,
    current_datetime as created_datetime,
    current_datetime as updated_datetime
  FROM `otherproject.dataset.sp.event_item_daily`
  WHERE file_ingested_date = load_dt);

  COMMIT TRANSACTION;

  --Roll back transaction
  EXCEPTION WHEN ERROR THEN
  ROLLBACK TRANSACTION;

  RAISE USING MESSAGE = FORMAT("Error Message - %s, Error Statement - %s, Error Stack - %s", @@error.message, @@error.statement_text, @@error.formatted_stack_trace);

END;

Tested by removing CASE statement with below and after that the creation of stored procedure succeeds.
if(epvid_vers_id like '%-%', null, `project.dataset.sp.udf_std_str_to_str`(epvid_vers_id)) as version_id

Version Details

JDBC Simba Driver Version: 1.3.3.1004
Liquibase BigQuery Version: 4.17.0

Output

Running Changeset: sp.sql::1::google
SEVERE [liquibase.changelog] ChangeSet sp.sql::1::google encountered an exception.
SEVERE [liquibase.integration] Migration failed for changeset sp.sql::1::gcp:
     Reason: liquibase.exception.DatabaseException: [Simba][BigQueryJDBCDriver](100032) Error executing query job. Message: EXEC_JOB_EXECUTION_ERR [Failed SQL: (100032) 

Steps to Reproduce

1. Create any Stored Procedure with CASE statement in it.
2. Run liquibase update

Expected Behavior

The stored procedure would've have been created.

Actual Behavior

Liquibase is throwing exception mentioned above.

Community Note

• Please vote on this issue by adding a 👍 reaction
  to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do
  not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment.

The generated sql for a "classic" add column is
ALTER TABLE <XXX> ADD <YYY> TYPE

For BigQuery the syntax is
ALTER TABLE ADD COLUMN TYPE

Community Note

• Please vote on this issue by adding a 👍 reaction
  to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do
  not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment.

Debug Output

https://gist.github.com/gabrielmdeal/85ba1470346861ef87779027556218ff

Also part of the gist is Array_Repro_Table.json, and the changelog.json that liquibase generateChangeLog creates.

Panic Output

N/A

Steps to Reproduce

TLDR: ARRAY and STRUCT column types don't seem to work with liquibase generateChangeLog. I can't find documentation explaining if it is possible to manually create JSON for these types of columns. Nor can I find documentation that says these types are unsupported.

1. Create multi-region datasets named Gabriel_Array and Gabriel_Array_Target
2. Get Array_Repro_Table.json from the gist above.
3. bq mk Gabriel_Array.Array_Repro_Table Array_Repro_Table.json
4. Run liquibase generateChangeLog on Gabriel_Array
5. Run liquibase update on Gabriel_Array_Target

I ran into a similar issue with a STRUCT column.

Expected Behavior

One of these:

• The Array_Repro_Table is created in Gabriel_Array_Target. 🎉
• liquibase generateChangeLog fails with an error saying that arrays and structs are not supported for Big Query.
• The documentation says that arrays and structs are not supported.

Actual Behavior

liquibase update fails with this error: liquibase.exception.DatabaseException: [Simba][BigQueryJDBCDriver](100032) Error executing query job. Message: Syntax error: Expected "<" but got "(" at [1:72] [Failed SQL: (100032) CREATE TABLE Gabriel_Array_Target.Array_Repro_Table (Array_Column ARRAY(65535))] See the gist for the full log.

Workaround

If I use --changeLogFile=changelog.bigquery.sql, then liquibase generateChangeLog still generates an invalid changelog.sql. But I am able to manually fix the SQL and then successfully run liquibase update.

Environment

Liquibase Version: 4.13.0 (latest now)

Liquibase Integration & Version: CLI

Liquibase Extension(s) & Version: liquibase-bigquery-4.13.0

Database Vendor & Version: Google BigQuery

Operating System Type & Version: Windows 10 21H2

Infrastructure Type/Provider: local

Description

Liquibase doesn't work properly with BigQuery, when dataset for DATABASECHANGELOG table is not in lowercase.

• PascalCase doesn't work (eg Changelog dataset, which is a valid name in BigQuery)
• underscores doesn't work (eg _changelog, which is a valid name too and useful to hide the dataset in BigQuery UI)

Most operations with DATABASECHANGELOG seems to work fine (changesets are applied and correctly excluded during the next liquibase update call). But there is one exception that breaks the CLI usage: when Liquibase tries to check if DATABASECHANGELOG table exists in the init phase - the code transforms dataset names to lowercase (Changeset -> changeset) and escapes underscores (_changelog -> \_changelog). Dataset names are case-sensitive in BigQuery, so JDBC driver returns null when DATABASECHANGELOG exists and Liquibase always tries to re-create this table, breaking the usage of liquibase update command.

• Lowercasing happens in https://github.com/liquibase/liquibase/blob/d2d634f7987a34f0ff1153a14a807a7bb02db611/liquibase-core/src/main/java/liquibase/database/AbstractJdbcDatabase.java#L318
• Escaping happens in https://github.com/liquibase/liquibase/blob/5273e472f74ba2df4434ad2ce9756a80e9865505/liquibase-core/src/main/java/liquibase/snapshot/JdbcDatabaseSnapshot.java#L1785

Found workarounds for lowercasing:

• Use lowercase for Liquibase dataset name (eg liquibaseSchemaName=changelog in configuration)
• Set liquibase.preserveSchemaCase=true property, but this requires an update to https://github.com/liquibase/liquibase-bigquery to set backtick (`) as escape character, because current extension uses double quotes ("), which breaks queries

It makes sense to update BigQuery extension documentation (https://docs.liquibase.com/install/tutorials/bigquery.html) with these limitations, so people don't stuck before BigQuery support is better.

I'm not sure how to fix support for underscores, as there is a private method in JdbcDatabaseSnapshot class that adds escaping.

Steps To Reproduce

1. Install Liquibase with BigQuery extension
2. Create any changelog, can be a simple SQL version with a single changeset

--liquibase formatted sql
--changeset author:1
CREATE SCHEMA Test;

3. Add Liquibase properties file with BigQuery configuration and PascalCase dataset name for Liquibase tables

changeLogFile=changelog.sql
liquibase.command.driver=com.simba.googlebigquery.jdbc.Driver
liquibase.command.url=jdbc:bigquery://https://googleapis.com/bigquery/v2:443;ProjectId={ProjectId};OAuthType=1;
liquibaseSchemaName=Changelog

4. Create Changelog dataset in your BigQuery project
5. Run liquibase update - it successfully creates DATABASECHANGELOG table in Changelog dataset and applies changesets
6. Run liquibase update again

Actual Behavior

Second update call fails with JDBC driver error, because it tries to create DATABASECHANGELOG table, that already exists. This can be inspected by liquibase update-sql call that displays CREATE TABLE DATABASECHANGELOG ... statement.

Expected/Desired Behavior

Second update call success and Liquibase doesn't try to re-create existing DATABASECHANGELOG table.

Hi,

I´m trying to execute some changelogs and I´ve de following error when liquidbase tries to create its tables:

[2023-11-23 10:47:49] INFO [liquibase.command] Command execution complete
[2023-11-23 10:47:49] SEVERE [liquibase.integration] [Simba]BigQueryJDBCDriver Error executing query job. Message: Type not found: VARCHAR at [1:36] [Failed SQL: (100032) CREATE TABLE DATABASECHANGELOG (ID VARCHAR(255) NOT NULL, AUTHOR VARCHAR(255) NOT NULL, FILENAME VARCHAR(255) NOT NULL, DATEEXECUTED datetime NOT NULL, ORDEREXECUTED INT NOT NULL, EXECTYPE VARCHAR(10) NOT NULL, MD5SUM VARCHAR(35), DESCRIPTION VARCHAR(255), COMMENTS VARCHAR(255), TAG VARCHAR(255), LIQUIBASE VARCHAR(20), CONTEXTS VARCHAR(255), LABELS VARCHAR(255), DEPLOYMENT_ID VARCHAR(10))]

• Caused by: 400 Bad Request
  POST https://bigquery.googleapis.com/bigquery/v2/projects/my-project/queries

"code": 400,
"errors": [
***
"domain": "global",
"location": "q",
"locationType": "parameter",
"message": "Type not found: VARCHAR at [1:36]",
"reason": "invalidQuery"
***
],
"message": "Type not found: VARCHAR at [1:36]",
"status": "INVALID_ARGUMENT"

liquibase.exception.CommandExecutionException: liquibase.exception.DatabaseException: liquibase.exception.DatabaseException: [Simba]BigQueryJDBCDriver Error executing query job. Message: Type not found: VARCHAR at [1:36] [Failed SQL: (100032) CREATE TABLE DATABASECHANGELOG (ID VARCHAR(255) NOT NULL, AUTHOR VARCHAR(255) NOT NULL, FILENAME VARCHAR(255) NOT NULL, DATEEXECUTED datetime NOT NULL, ORDEREXECUTED INT NOT NULL, EXECTYPE VARCHAR(10) NOT NULL, MD5SUM VARCHAR(35), DESCRIPTION VARCHAR(255), COMMENTS VARCHAR(255), TAG VARCHAR(255), LIQUIBASE VARCHAR(20), CONTEXTS VARCHAR(255), LABELS VARCHAR(255), DEPLOYMENT_ID VARCHAR(10))]
at liquibase.command.CommandScope.execute(CommandScope.java:237)
at liquibase.integration.commandline.CommandRunner.call(CommandRunner.java:55)
at liquibase.integration.commandline.CommandRunner.call(CommandRunner.java:24)
at picocli.CommandLine.executeUserObject(CommandLine.java:2041)
at picocli.CommandLine.access$1500(CommandLine.java:148)
at picocli.CommandLine$RunLast.executeUserObjectOfLastSubcommandWithSameParent(CommandLine.java:2461)
at picocli.CommandLine$RunLast.handle(CommandLine.java:2453)
at picocli.CommandLine$RunLast.handle(CommandLine.java:2415)
at picocli.CommandLine$AbstractParseResultHandler.execute(CommandLine.java:2273)
at picocli.CommandLine$RunLast.execute(CommandLine.java:2417)
at picocli.CommandLine.execute(CommandLine.java:2170)

I´m using the liquidbase github action plugin with the following options:

• uses: liquibase-github-actions/[email protected]
  id: updatedb
  name: Update liquibase
  with:
  changelogFile: "db-changelog/config/changelog-master.xml"
  url: "jdbc:bigquery://https://www.googleapis.com/bigquery/v2:443;OAuthType=3;ProjectId=my-project;EnableSession=1;DefaultDataset=test_github;Location=europe-west1;QueryDialect=SQL"
  driver: "com.simba.googlebigquery.jdbc42.Driver"

is it maybe liquidbase library github action plugin is using?

Many thanks

Caused by: java.sql.SQLException: [Simba]BigQueryJDBCDriver Error executing query job. Message: Type not found: VARCHAR at [1:41]

should be STRING

Caused by: liquibase.exception.DatabaseException: [Simba]BigQueryJDBCDriver Error executing query job. Message: Type not found: VARCHAR at [1:41] [Failed SQL: (100032) CREATE TABLE mydb.DATABASECHANGELOG (ID VARCHAR(255) NOT NULL, AUTHOR VARCHAR(255) NOT NULL, FILENAME VARCHAR(255) NOT NULL, DATEEXECUTED datetime NOT NULL, ORDEREXECUTED INT NOT NULL, EXECTYPE VARCHAR(10) NOT NULL, MD5SUM VARCHAR(35), DESCRIPTION VARCHAR(255), COMMENTS VARCHAR(255), TAG VARCHAR(255), LIQUIBASE VARCHAR(20), CONTEXTS VARCHAR(255), LABELS VARCHAR(255), DEPLOYMENT_ID VARCHAR(10))]

Issues on GitHub are intended to be related to bugs or feature requests with provider codebase,
so we recommend using our other community resources instead of asking here 👍.

I am trying to run the following command on liquibase 4.21.1 and the liquibase-bigquery extension is also 4.21.1.

liquibase update --changelog-file="/db/changelog/changelog-master.yaml" --url="jdbc:bigquery://https://googleapis.com/bigquery/v2:443/my_bq_db;ProjectId=my-gcp-project;OAuthType=0;[email protected];OAuthPvtKeyPath=/gcloud/my-gcp-project-61e489b85654.json" --driver="com.simba.googlebigquery.jdbc42.Driver" --classpath="/SimbaJDBCDriverforGoogleBigQuery42_1.3.3.1004/GoogleBigQueryJDBC42.jar"

this is complaining with the following error:

Unexpected error running Liquibase: [Simba]BigQueryJDBCDriver Error executing query job. Message: Table "DATABASECHANGELOGLOCK" must be qualified with a dataset (e.g. dataset.table). [Failed SQL: (100032) CREATE TABLE DATABASECHANGELOGLOCK (ID INT, LOCKED BOOLEAN, LOCKGRANTED datetime, LOCKEDBY STRING(255))]

• Caused by: 400 Bad Request
  POST https://bigquery.googleapis.com/bigquery/v2/projects/my-gcp-project/queries
  {
  "code": 400,
  "errors": [
  {
  "domain": "global",
  "message": "Table "DATABASECHANGELOGLOCK" must be qualified with a dataset (e.g. dataset.table).",
  "reason": "invalid"
  }
  ],
  "message": "Table "DATABASECHANGELOGLOCK" must be qualified with a dataset (e.g. dataset.table).",
  "status": "INVALID_ARGUMENT"
  }

I was following the following page: https://contribute.liquibase.com/extensions-integrations/directory/database-tutorials/bigquery/

Was wondering if there's a workaround for this?

Community Note

• Please vote on this issue by adding a 👍 reaction
  to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do
  not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment.

Debug Output

N/A

Panic Output

N/A

Steps to Reproduce

1. Use Liquibase v4.27.0.
2. Have a dataset set to Multi-region EU.
3. Have a dataset replica set up in the US.
4. Do not specify the Location in the URL
5. Run a Liquibase command

Expected Behavior

The expectation is that it would connect to the EU location without needing to specify it in the URL.

Actual Behavior

Liquibase is connecting to the US region even though is it the secondary replica which is read-only.

{
  "code": 400,
  "errors": [
    {
      "domain": "global",
      "message": "The dataset replica of the cross region dataset '<host>:<dataset>' in region 'US' is read-only because it's not the primary replica.",
      "reason": "invalid"
    }
  ],
  "message": "The dataset replica of the cross region dataset '<host>:<dataset>' in region 'US' is read-only because it's not the primary replica.",
  "status": "INVALID_ARGUMENT"
}

References

• #0000

Users are not currently able to use generateChangeLog successfull

Community Note

• Please vote on this issue by adding a 👍 reaction
  to the original issue to help the community and maintainers prioritize this request
• Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do
  not help prioritize the request
• If you are interested in working on this issue or have submitted a pull request, please leave a comment.

Debug Output

https://gist.github.com/gabrielmdeal/f82a75154fefbade72061d95c6d6fd2d

The gist also contains the Repro_Table.json and Repro_View.sql files that are referenced below.

Panic Output

N/A

Steps to Reproduce

TLDR: liquibase generateChangeLog fails if my dataset is regional and contains a view. It works for a multi-region dataset.

1. Install:
   • The liquibase/liquibase:4.22 image from hub.docker.com
   • https://storage.googleapis.com/simba-bq-release/jdbc/SimbaBigQueryJDBC42-1.3.2.1003.zip
   • https://github.com/liquibase/liquibase-bigquery/releases/download/v4.22.0/liquibase-bigquery-4.22.0.jar
2. Create a Big Query dataset named Gabriel_Liquibase in the us-central1 region. Not multi-region.
3. Create a liquibase service account with the BigQuery Admin, BigQuery Data Owner and BigQuery Metadata Viewer roles. (I know this is overkill, but I knee-jerk added more roles when I ran into this error.)
4. gcloud iam service-accounts keys create "liquibase-service-account-credentials.json" --iam-account "[email protected]"
5. Use the JSON and SQL from the above gist to create a table and a view:

   bq mk Gabriel_Liquibase.Repro_Table Repro_Table.json
   bq mk --use_legacy_sql=false --view "$(cat Repro_View.sql)" Gabriel_Liquibase.Repro_View
   

6. Try to generate a change log via liquibase generateChangeLog --url 'jdbc:bigquery://https://www.googleapis.com/bigquery/v2:443;ProjectId=redacted-project-id;DefaultDataset=Gabriel_Liquibase;OAuthType=0;OAuthServiceAcctEmail=liquibase@redacted-project-id.iam.gserviceaccount.com;OAuthPvtKeyPath=liquibase-service-account-credentials.json' --logLevel DEBUG --changeLogFile changelog.json --driver com.simba.googlebigquery.jdbc.Driver

Expected Behavior

A changelog is created with my table and view.

Actual Behavior

The changelog is not created. liquibase generateChangeLog fails with the below error (see gist for full output). If I use console.cloud.google.com to run the SQL from the error, I do get the expected view definition.

The liquibase generateChangeLog command succeeds if I delete my view.

[2023-05-16 21:39:35] SEVERE [liquibase.integration] Error executing SQL SELECT view_definition FROM Gabriel_Liquibase.INFORMATION_SCHEMA.VIEWS WHERE table_name='Repro_View' AND table_schema='Gabriel_Liquibase' AND table_catalog='redacted-project-id';: [Simba][BigQueryJDBCDriver](100032) Error executing query job. Message: Not found: Dataset redacted-project-id:Gabriel_Liquibase was not found in location US
  - Caused by: 404 Not Found
POST https://bigquery.googleapis.com/bigquery/v2/projects/redacted-project-id/queries
{
  "code": 404,
  "errors": [
    {
      "domain": "global",
      "message": "Not found: Dataset redacted-project-id:Gabriel_Liquibase was not found in location US",
      "reason": "notFound"
    }
  ],
  "message": "Not found: Dataset redacted-project-id:Gabriel_Liquibase was not found in location US",
  "status": "NOT_FOUND"
}

Need to more those test cases as valid and get them working in LTH.

Dockerfile

FROM liquibase/liquibase:4.13.0
USER 0:0
COPY liquibase.docker.properties /liquibase/
COPY key.json /liquibase/
COPY changelog/ /liquibase/changelog/
RUN apt update && apt install curl unzip -y
RUN curl -SsL https://storage.googleapis.com/simba-bq-release/jdbc/SimbaJDBCDriverforGoogleBigQuery42_1.2.23.1027.zip -o SimbaJDBCDriverforGoogleBigQuery.zip \
 && unzip SimbaJDBCDriverforGoogleBigQuery.zip -d /liquibase/lib/ \
 && curl -SsL https://github.com/liquibase/liquibase-bigquery/releases/download/liquibase-bigquery-4.13.0/liquibase-bigquery-4.13.0.jar -o /liquibase/lib/liquibase-bigquery-4.13.0.jar

liquibase.docker.properties

classpath: /liquibase/changelog
url: jdbc:bigquery://https://www.googleapis.com/bigquery/v2:443;ProjectId=[ProjectId];OAuthType=0;OAuthServiceAcctEmail=[ServiceAccount];OAuthPvtKeyPath=/liquibase/key.json
changeLogFile: changelog.sql
username:
password:
databaseChangeLogLockTableName: [ProjectId].[DataSetName].databasechangeloglock
databaseChangeLogTableName: [ProjectId].[DataSetName].databasechangelog

changelog/changelog.sql

-- liquibase formatted sql

-- changeset liquibase:1
CREATE TABLE test_table (test_id INT, test_column VARCHAR, PRIMARY KEY (test_id))

Then execute this command.

docker build -t liquibase:0.0.1 . --no-cache=true
docker run --rm liquibase:0.0.1 update --log-level=fine

Then an error occurs.

Caused by: liquibase.exception.DatabaseException: [Simba][BigQueryJDBCDriver](100032) Error executing query job. Message: Syntax error: Unexpected string literal "[ProjectId].[DataSetName]...." at [1:14] [Failed SQL: (100032) CREATE TABLE "[ProjectId].[DataSetName].databasechangeloglock" (ID INT NOT NULL, LOCKED BOOLEAN NOT NULL, LOCKGRANTED datetime, LOCKEDBY STRING(255))]

I get an error when creating the DATABASECHANGELOGLOCK table.
The cause of the error seems to be that the TABLE name is enclosed in double quotes.
Why do I get this error?