🎉 Refactor Normalization docker images and upgrade to use dbt 0.21.0 (a…

…irbytehq#6959)

* Split normalization docker images for some connectors with specifics dependencies

* Regenerate (airbytehq#7003)

airbyte-integrations/bases/base-normalization/.dockerignore

@@ -5,3 +5,6 @@
 !setup.py
 !normalization
 !dbt-project-template
+!dbt-project-template-mssql
+!dbt-project-template-mysql
+!dbt-project-template-oracle

airbyte-integrations/bases/base-normalization/Dockerfile

@@ -1,36 +1,4 @@
-FROM fishtownanalytics/dbt:0.19.0
-
-USER root
-WORKDIR /tmp
-RUN apt-get update && apt-get install -y \
-    wget \
-    curl \
-    unzip \
-    libaio-dev \
-    libaio1 \
-    gnupg \
-    gnupg1 \
-    gnupg2
-
-# Install MS SQL Server dependencies
-RUN curl https://packages.microsoft.com/keys/microsoft.asc | apt-key add -
-RUN curl https://packages.microsoft.com/config/debian/10/prod.list > /etc/apt/sources.list.d/mssql-release.list
-RUN apt-get update && ACCEPT_EULA=Y apt-get install -y \
-    libgssapi-krb5-2 \
-    unixodbc-dev \
-    msodbcsql17 \
-    mssql-tools
-ENV PATH=$PATH:/opt/mssql-tools/bin
-
-# Install Oracle dependencies
-RUN mkdir -p /opt/oracle
-RUN wget https://download.oracle.com/otn_software/linux/instantclient/19600/instantclient-basic-linux.x64-19.6.0.0.0dbru.zip
-RUN unzip instantclient-basic-linux.x64-19.6.0.0.0dbru.zip -d /opt/oracle
-ENV ORACLE_HOME /opt/oracle/instantclient_19_6
-ENV LD_LIBRARY_PATH=$LD_LIBRARY_PATH:$ORACLE_HOME
-ENV TNS_ADMIN /opt/oracle/instantclient_19_6/network/admin
-RUN pip install cx_Oracle
-
+FROM fishtownanalytics/dbt:0.21.0
 COPY --from=airbyte/base-airbyte-protocol-python:0.1.1 /airbyte /airbyte
 
 # Install SSH Tunneling dependencies
@@ -50,10 +18,6 @@ RUN pip install .
 
 WORKDIR /airbyte/normalization_code
 RUN pip install .
-RUN pip install dbt-oracle==0.4.3
-RUN pip install git+https://github.com/dbeatty10/dbt-mysql@96655ea9f7fca7be90c9112ce8ffbb5aac1d3716#egg=dbt-mysql
-RUN pip install dbt-sqlserver==0.19.3
-
 
 WORKDIR /airbyte/normalization_code/dbt-template/
 # Download external dbt dependencies
@@ -63,5 +27,4 @@ WORKDIR /airbyte
 ENV AIRBYTE_ENTRYPOINT "/airbyte/entrypoint.sh"
 ENTRYPOINT ["/airbyte/entrypoint.sh"]
 
-LABEL io.airbyte.version=0.1.52
 LABEL io.airbyte.name=airbyte/normalization

airbyte-integrations/bases/base-normalization/README.md

@@ -108,13 +108,27 @@ or can also be invoked on github, thanks to the slash commands posted as comment
 
     /test connector=bases/base-normalization
 
+You can restrict the tests to a subset of destinations by specifying a comma separated list of destinations.
+For example, let's say you are working on a change to normalization for Postgres, with Gradle:
+
+    NORMALIZATION_TEST_TARGET=postgres ./gradlew :airbyte-integrations:bases:base-normalization:integrationTest
+
+or directly with pytest:
+
+    NORMALIZATION_TEST_TARGET=postgres  pytest airbyte-integrations/bases/base-normalization/integration_tests
+
 Note that these tests are connecting and processing data on top of real data warehouse destinations.
 Therefore, valid credentials files are expected to be injected in the `secrets/` folder in order to run 
 (not included in git repository).
 
 This is usually automatically done by the CI thanks to the `tools/bin/ci_credentials.sh` script or you can 
 re-use the `destination_config.json` passed to destination connectors.
 
+As normalization supports more and more destinations, tests are relying on an increasing number of destinations. 
+As a result, it is possible that the docker garbage collector is triggered to wipe "unused" docker images while the 
+integration tests for normalization are running. Thus, if you encounter errors about a connector's docker image not being 
+present locally (even though it was built beforehand), make sure to increase the docker image storage size of your docker engine ("defaultKeepStorage" for mac for example).
+
 ### Integration Tests Definitions for test_ephemeral.py:
 The test here focus on benchmarking the "ephemeral" materialization mode of dbt. Depending on the number of
 columns in a catalog, this may throw exceptions and fail. This test ensures that we support reasonable number of columns in destination tables.

airbyte-integrations/bases/base-normalization/build.gradle

@@ -1,3 +1,5 @@
+import java.nio.file.Paths
+
 plugins {
     id 'airbyte-docker'
     id 'airbyte-python'
@@ -27,13 +29,52 @@ task checkSshScriptCopy(type: Task, dependsOn: copySshScript) {
     }
 }
 
-test.dependsOn checkSshScriptCopy
-assemble.dependsOn checkSshScriptCopy
+airbyteDocker.dependsOn(checkSshScriptCopy)
+assemble.dependsOn(checkSshScriptCopy)
+test.dependsOn(checkSshScriptCopy)
 
 installReqs.dependsOn(":airbyte-integrations:bases:airbyte-protocol:installReqs")
 integrationTest.dependsOn(build)
 
-task("customIntegrationTestPython", type: PythonTask, dependsOn: installTestReqs){
+
+static def getDockerfile(String customConnector) {
+    return "${customConnector}.Dockerfile"
+}
+
+static def getDockerImageName(String customConnector) {
+    return "airbyte/normalization-${customConnector}"
+}
+
+static def getImageNameWithTag(String customConnector) {
+    return "${getDockerImageName(customConnector)}:dev"
+}
+
+
+def buildAirbyteDocker(String customConnector) {
+    def baseCommand = ['docker', 'build', '.', '-f', getDockerfile(customConnector), '-t', getImageNameWithTag(customConnector)]
+    return {
+        commandLine baseCommand
+    }
+}
+
+task airbyteDockerMSSql(type: Exec, dependsOn: checkSshScriptCopy) {
+    configure buildAirbyteDocker('mssql')
+    dependsOn assemble
+}
+task airbyteDockerMySql(type: Exec, dependsOn: checkSshScriptCopy) {
+    configure buildAirbyteDocker('mysql')
+    dependsOn assemble
+}
+task airbyteDockerOracle(type: Exec, dependsOn: checkSshScriptCopy) {
+    configure buildAirbyteDocker('oracle')
+    dependsOn assemble
+}
+
+airbyteDocker.dependsOn(airbyteDockerMSSql)
+airbyteDocker.dependsOn(airbyteDockerMySql)
+airbyteDocker.dependsOn(airbyteDockerOracle)
+
+task("customIntegrationTestPython", type: PythonTask, dependsOn: installTestReqs) {
     module = "pytest"
     command = "-s integration_tests"
 
@@ -45,7 +86,6 @@ task("customIntegrationTestPython", type: PythonTask, dependsOn: installTestReqs
     dependsOn ':airbyte-integrations:connectors:destination-snowflake:airbyteDocker'
     dependsOn ':airbyte-integrations:connectors:destination-oracle:airbyteDocker'
     dependsOn ':airbyte-integrations:connectors:destination-mssql:airbyteDocker'
-
 }
 
 integrationTest.dependsOn("customIntegrationTestPython")

airbyte-integrations/bases/base-normalization/dbt-project-template-mssql/dbt_project.yml

@@ -0,0 +1,58 @@
+# This file is necessary to install dbt-utils with dbt deps
+# the content will be overwritten by the transform function
+
+# Name your package! Package names should contain only lowercase characters
+# and underscores. A good package name should reflect your organization's
+# name or the intended use of these models
+name: 'airbyte_utils'
+version: '1.0'
+config-version: 2
+
+# This setting configures which "profile" dbt uses for this project. Profiles contain
+# database connection information, and should be configured in the  ~/.dbt/profiles.yml file
+profile: 'normalize'
+
+# These configurations specify where dbt should look for different types of files.
+# The `source-paths` config, for example, states that source models can be found
+# in the "models/" directory. You probably won't need to change these!
+source-paths: ["models"]
+docs-paths: ["docs"]
+analysis-paths: ["analysis"]
+test-paths: ["tests"]
+data-paths: ["data"]
+macro-paths: ["macros"]
+
+target-path: "../build"  # directory which will store compiled SQL files
+log-path: "../logs"  # directory which will store DBT logs
+modules-path: "/tmp/dbt_modules"  # directory which will store external DBT dependencies
+
+clean-targets:         # directories to be removed by `dbt clean`
+    - "build"
+    - "dbt_modules"
+
+quoting:
+  database: true
+# Temporarily disabling the behavior of the ExtendedNameTransformer on table/schema names, see (issue #1785)
+# all schemas should be unquoted
+  schema: false
+  identifier: true
+
+# You can define configurations for models in the `source-paths` directory here.
+# Using these configurations, you can enable or disable models, change how they
+# are materialized, and more!
+models:
+  airbyte_utils:
+    generated:
+      airbyte_ctes:
+        +tags: airbyte_internal_cte
+        +materialized: ephemeral
+      airbyte_views:
+        +tags: airbyte_internal_views
+        +materialized: view
+      airbyte_tables:
+        +tags: normalized_tables
+        +materialized: table
+    +materialized: table
+
+vars:
+  dbt_utils_dispatch_list: ['airbyte_utils']

airbyte-integrations/bases/base-normalization/dbt-project-template-mssql/packages.yml

@@ -0,0 +1,5 @@
+# add dependencies. these will get pulled during the `dbt deps` process.
+
+packages:
+  - git: "https://github.com/fishtown-analytics/dbt-utils.git"
+    revision: 0.6.4

airbyte-integrations/bases/base-normalization/dbt-project-template-mysql/dbt_project.yml

@@ -0,0 +1,58 @@
+# This file is necessary to install dbt-utils with dbt deps
+# the content will be overwritten by the transform function
+
+# Name your package! Package names should contain only lowercase characters
+# and underscores. A good package name should reflect your organization's
+# name or the intended use of these models
+name: 'airbyte_utils'
+version: '1.0'
+config-version: 2
+
+# This setting configures which "profile" dbt uses for this project. Profiles contain
+# database connection information, and should be configured in the  ~/.dbt/profiles.yml file
+profile: 'normalize'
+
+# These configurations specify where dbt should look for different types of files.
+# The `source-paths` config, for example, states that source models can be found
+# in the "models/" directory. You probably won't need to change these!
+source-paths: ["models"]
+docs-paths: ["docs"]
+analysis-paths: ["analysis"]
+test-paths: ["tests"]
+data-paths: ["data"]
+macro-paths: ["macros"]
+
+target-path: "../build"  # directory which will store compiled SQL files
+log-path: "../logs"  # directory which will store DBT logs
+modules-path: "/tmp/dbt_modules"  # directory which will store external DBT dependencies
+
+clean-targets:         # directories to be removed by `dbt clean`
+    - "build"
+    - "dbt_modules"
+
+quoting:
+  database: true
+# Temporarily disabling the behavior of the ExtendedNameTransformer on table/schema names, see (issue #1785)
+# all schemas should be unquoted
+  schema: false
+  identifier: true
+
+# You can define configurations for models in the `source-paths` directory here.
+# Using these configurations, you can enable or disable models, change how they
+# are materialized, and more!
+models:
+  airbyte_utils:
+    generated:
+      airbyte_ctes:
+        +tags: airbyte_internal_cte
+        +materialized: ephemeral
+      airbyte_views:
+        +tags: airbyte_internal_views
+        +materialized: view
+      airbyte_tables:
+        +tags: normalized_tables
+        +materialized: table
+    +materialized: table
+
+vars:
+  dbt_utils_dispatch_list: ['airbyte_utils']

airbyte-integrations/bases/base-normalization/dbt-project-template-mysql/packages.yml

@@ -0,0 +1,5 @@
+# add dependencies. these will get pulled during the `dbt deps` process.
+
+packages:
+  - git: "https://github.com/fishtown-analytics/dbt-utils.git"
+    revision: 0.6.4

airbyte-integrations/bases/base-normalization/dbt-project-template-oracle/dbt_project.yml

@@ -0,0 +1,56 @@
+# This file is necessary to install dbt-utils with dbt deps
+# the content will be overwritten by the transform function
+
+# Name your package! Package names should contain only lowercase characters
+# and underscores. A good package name should reflect your organization's
+# name or the intended use of these models
+name: 'airbyte_utils'
+version: '1.0'
+config-version: 2
+
+# This setting configures which "profile" dbt uses for this project. Profiles contain
+# database connection information, and should be configured in the  ~/.dbt/profiles.yml file
+profile: 'normalize'
+
+# These configurations specify where dbt should look for different types of files.
+# The `source-paths` config, for example, states that source models can be found
+# in the "models/" directory. You probably won't need to change these!
+source-paths: ["models"]
+docs-paths: ["docs"]
+analysis-paths: ["analysis"]
+test-paths: ["tests"]
+data-paths: ["data"]
+macro-paths: ["macros"]
+
+target-path: "../build"  # directory which will store compiled SQL files
+log-path: "../logs"  # directory which will store DBT logs
+modules-path: "/tmp/dbt_modules"  # directory which will store external DBT dependencies
+
+clean-targets:         # directories to be removed by `dbt clean`
+    - "build"
+    - "dbt_modules"
+
+quoting:
+  database: false
+  schema: false
+  identifier: false
+
+# You can define configurations for models in the `source-paths` directory here.
+# Using these configurations, you can enable or disable models, change how they
+# are materialized, and more!
+models:
+  airbyte_utils:
+    generated:
+      airbyte_ctes:
+        +tags: airbyte_internal_cte
+        +materialized: ephemeral
+      airbyte_views:
+        +tags: airbyte_internal_views
+        +materialized: view
+      airbyte_tables:
+        +tags: normalized_tables
+        +materialized: table
+    +materialized: table
+
+vars:
+  dbt_utils_dispatch_list: ['airbyte_utils']

airbyte-integrations/bases/base-normalization/dbt-project-template-oracle/packages.yml

@@ -0,0 +1,5 @@
+# add dependencies. these will get pulled during the `dbt deps` process.
+
+packages:
+  - git: "https://github.com/fishtown-analytics/dbt-utils.git"
+    revision: 0.6.4

airbyte-integrations/bases/base-normalization/dbt-project-template/dbt_project.yml

@@ -54,5 +54,6 @@ models:
         +materialized: table
     +materialized: table
 
-vars:
-  dbt_utils_dispatch_list: ['airbyte_utils']
+dispatch:
+  - macro_namespace: dbt_utils
+    search_order: ['airbyte_utils', 'dbt_utils']

airbyte-integrations/bases/base-normalization/dbt-project-template/macros/cross_db_utils/concat.sql

@@ -1,19 +1,16 @@
 {#
-    Overriding the following macro from dbt-utils:
-        https://github.com/fishtown-analytics/dbt-utils/blob/0.6.2/macros/cross_db_utils/concat.sql
-    To implement our own version of concat
-    Because on postgres, we cannot pass more than 100 arguments to a function
-    This is necessary until: https://github.com/fishtown-analytics/dbt-utils/blob/dev/0.7.0/macros/cross_db_utils/concat.sql
-    is released.
+    concat in dbt 0.6.4 used to work fine for bigquery but the new implementaion in 0.7.3 is less scalable (can not handle too many columns)
+    Therefore, we revert the implementation here and add versions for missing destinations
 #}
 
 {% macro concat(fields) -%}
-  {{ adapter.dispatch('concat', packages = ['airbyte_utils', 'dbt_utils'])(fields) }}
+  {{ adapter.dispatch('concat')(fields) }}
 {%- endmacro %}
 
-{% macro postgres__concat(fields) %}
-    {{ dbt_utils.alternative_concat(fields) }}
-{% endmacro %}
+{% macro bigquery__concat(fields) -%}
+    {#-- concat() in SQL bigquery scales better with number of columns than using the '||' operator --#}
+    concat({{ fields|join(', ') }})
+{%- endmacro %}
 
 {% macro sqlserver__concat(fields) -%}
     {#-- CONCAT() in SQL SERVER accepts from 2 to 254 arguments, we use batches for the main concat, to overcome the limit. --#}

airbyte-integrations/bases/base-normalization/dbt-project-template/macros/cross_db_utils/drop_schema.sql

@@ -0,0 +1,8 @@
+{#
+    Drop schema to clean up the destination database
+#}
+{% macro drop_schemas(schemas) %}
+  {% for schema in schemas %}
+    drop schema if exists {{ schema }} cascade;
+  {% endfor %}
+{% endmacro %}