DRILL-5978: Updating of Apache and MapR Hive libraries to 2.3.2 and 2.1.2-mapr-1710 versions respectively by vdiravka · Pull Request #1111 · apache/drill

vdiravka · 2018-02-05T19:41:04Z

Updating hive properties for tests and resolving dependencies and API conflicts:

Allowing of using Hive's own calcite-core and avatica versions by hive-exec.
Calcite version is removed from root POM Dependency Management
Fix for "hive.metastore.schema.verification", MetaException(message:
Version information not found in metastore)
https://cwiki.apache.org/confluence/display/Hive/Hive+Schema+Tool
METASTORE_SCHEMA_VERIFICATION="false" property is added
Fix JSONException class is not found (excluded banned org.json dependency)
Added METASTORE_AUTO_CREATE_ALL="true", properties to tests, because some additional
tables are necessary in Hive metastore
Disabling calcite CBO for (Hive's CalcitePlanner) for tests, because it is in conflict
with Drill's Calcite version for Drill unit tests. HIVE_CBO_ENABLED="false" property
jackson and parquet libraries are relocated in hive-exec-shade module
org.apache.parquet:parquet-column Drill version is added to "hive-exec" to
allow of using Parquet empty group on MessageType level (PARQUET-278)
Removing of commons-codec exclusion from hive core. This dependency is
necessary for hive-exec and hive-metastore.
Setting Hive's internal properties for transactional scan:
HiveConf.HIVE_TRANSACTIONAL_TABLE_SCAN and for schema evolution: HiveConf.HIVE_SCHEMA_EVOLUTION,
IOConstants.SCHEMA_EVOLUTION_COLUMNS, IOConstants.SCHEMA_EVOLUTION_COLUMNS_TYPES

priteshm · 2018-02-05T20:04:23Z

@vrozov or @sohami can you please review this change?

@vdiravka can you also add the Drill JIRA in the title?

sohami · 2018-02-06T06:42:09Z

          </exclusion>
+          <exclusion>
+            <groupId>org.mortbay.jetty</groupId>
+            <artifactId>servlet-api</artifactId>


duplicate exclusion. Already excluded at the top

You are right. Exclusion is removed. Thank you.

sohami · 2018-02-06T07:07:13Z

    hiveConfig.put(METASTORE_EXECUTE_SET_UGI.varname, hiveConf.get(METASTORE_EXECUTE_SET_UGI.varname));
    hiveConfig.put(HIVE_AUTHORIZATION_ENABLED.varname, hiveConf.get(HIVE_AUTHORIZATION_ENABLED.varname));
    hiveConfig.put(HIVE_AUTHENTICATOR_MANAGER.varname, SessionStateUserAuthenticator.class.getName());
-    hiveConfig.put(HIVE_AUTHORIZATION_MANAGER.varname, SQLStdHiveAuthorizerFactory.class.getName());


why is this removed ? The test seem to be for Authorization.

Did it accidentally. I've returned this string.

sohami · 2018-02-06T07:50:33Z

+          if (AcidUtils.isTablePropertyTransactional(properties)) {
+            AcidUtils.setTransactionalTableScan(job, true);
+            HiveUtilities.setColumnTypes(job, properties, true, sd);
+          }


How about refactoring this block of code to a new method in HiveUtilities ? Like verifyAndAddTransactionalProperty(). Then just call that method from both here and HiveAbstractReader

It makes sense. Moreover since schema_evolution is required for acid tables HIVE-12799 I've combined it with setColumnTypes() helper method.

sohami · 2018-02-06T08:16:01Z

+        colTypesBuilder.append(col.getType());
+      }
+      colNames = colNamesBuilder.toString();
+      colTypes = colTypesBuilder.toString();


how about changing the loop as below:

final StringBuilder colNamesBuilder = new StringBuilder(); final StringBuilder colTypesBuilder = new StringBuilder(); for(FieldSchema col: sd.getCols()) { colNamesBuilder.append(col.getName()); colTypesBuilder.append(col.getType()); colNamesBuilder.append(','); colTypesBuilder.append(','); } colNames = colNamesBuilder.substring(0, colNamesBuilder.length() - 1); colTypes = colTypesBuilder.substring(0, colTypesBuilder.length() - 1);

I like it. Thank you.

vdiravka

@sohami Thank you for CR. All changes were done.

vdiravka · 2018-02-06T10:19:47Z

    hiveConfig.put(METASTORE_EXECUTE_SET_UGI.varname, hiveConf.get(METASTORE_EXECUTE_SET_UGI.varname));
    hiveConfig.put(HIVE_AUTHORIZATION_ENABLED.varname, hiveConf.get(HIVE_AUTHORIZATION_ENABLED.varname));
    hiveConfig.put(HIVE_AUTHENTICATOR_MANAGER.varname, SessionStateUserAuthenticator.class.getName());
-    hiveConfig.put(HIVE_AUTHORIZATION_MANAGER.varname, SQLStdHiveAuthorizerFactory.class.getName());


Did it accidentally. I've returned this string.

vdiravka · 2018-02-06T10:22:30Z

          </exclusion>
+          <exclusion>
+            <groupId>org.mortbay.jetty</groupId>
+            <artifactId>servlet-api</artifactId>


You are right. Exclusion is removed. Thank you.

vdiravka · 2018-02-06T10:34:53Z

+        colTypesBuilder.append(col.getType());
+      }
+      colNames = colNamesBuilder.toString();
+      colTypes = colTypesBuilder.toString();


I like it. Thank you.

vdiravka · 2018-02-06T12:37:07Z

+          if (AcidUtils.isTablePropertyTransactional(properties)) {
+            AcidUtils.setTransactionalTableScan(job, true);
+            HiveUtilities.setColumnTypes(job, properties, true, sd);
+          }


It makes sense. Moreover since schema_evolution is required for acid tables HIVE-12799 I've combined it with setColumnTypes() helper method.

vrozov · 2018-02-07T03:03:06Z

    <dependency>
      <groupId>org.apache.calcite</groupId>
      <artifactId>calcite-core</artifactId>
+      <version>${calcite.version}</version>


Why is this change necessary? The version should come from <dependencyManagement> of the parent pom.

hive-exec needs own calcite 1.6 version.
Calcite version in DependencyManagement of root pom leads to using Drill Calcite version over the whole project.

Will it be better to explicitly specify the version of calcite-core in hive-exec?

I have returned calcite-core version into DependencyManagement block. Drill Calcite version libraries are included into "drill-hive-exec-shaded" module.

I will tell you why it works for now:
When user submits query in Drill via Hive plugin, the query is validated and planned via Drill Calcite. So Hive Calcite isn't necessary for it.
Hive Calcite is used only in the process of Drill unit testing, where a lot of Hive specific queries are performed to setup Hive store for testing. But Drill Calcite and Avatica versions have conflicts with Hive old Calcite and Avatica versions. That's why I have disabled Calcite cost based optimizator conf.set(ConfVars.HIVE_CBO_ENABLED.varname, "false");. We can enable it again once Hive will leverage the newest Calcite version. The comment about it is added into drill-hive-exec-shaded POM.

vrozov · 2018-02-07T03:05:03Z

    <dependency>
      <groupId>org.apache.calcite</groupId>
      <artifactId>calcite-core</artifactId>
+      <version>${calcite.version}</version>


The same question as for common/pom.xml.

vrozov · 2018-02-07T03:07:47Z

          <groupId>commons-codec</groupId>
          <artifactId>commons-codec</artifactId>
        </exclusion>
+        <exclusion>


Is the exclusion necessary due to a version conflict or the dependency is not required?

version conflict.

For a transitive dependency will it be better to use DependencyManagement? Please add comments with details.

Both cases can resolve it. In details io.dropwizard.metrics:metrics-core is nowhere used in Drill. And this is a dependency for tephra-core and transitive for hive-metastore. But it conflicts with Drill's com.codahale.metrics.
hive-metastore uses 3.0.1 version of this dependency, but the last version in maven repository is 4.0.2.
I added this dependency to dependencyManagement block with 4.0.2 version and conflict is resolved as well. I think it is a better decision, because it can help to avoid similar conflicts in future.

Also metrics-core in hive-hbase-handler has not influence to Drill, so I've removed my exclusion of it.

vrozov · 2018-02-07T03:26:03Z

+   * @param properties table or partition properties
+   * @param sd storage descriptor
+   */
+  public static void verifyAndAddTransactionalProperties(JobConf job, Properties properties, StorageDescriptor sd) {


Is it necessary to pass both JobConf and Properties? As far as I can see job is always populated using passed properties.

job involves table properties, but JobConf hasn't any method to return that properties. It is possible to get only one property. But looks like this is not an issue.
There is in AcidUtils the isTablePropertyTransactional method with Configuration input parameter.
Changed. Thank you.

vrozov · 2018-02-07T03:43:48Z

+        colNames = job.get(serdeConstants.LIST_COLUMNS);
+        colTypes = job.get(serdeConstants.LIST_COLUMN_TYPES);
+      } else {
+        final StringBuilder colNamesBuilder = new StringBuilder();


consider using Joiner.

Thank you, it's cleaner.

vrozov · 2018-02-07T03:45:24Z

+
+      // Try to get get column names and types from table or partition properties. If they are absent there, get columns
+      // data from storage descriptor of the table
+      if (properties.containsKey(serdeConstants.LIST_COLUMNS) && properties.containsKey(serdeConstants.LIST_COLUMN_TYPES)) {


avoid double get() (containsKey()) if possible.

Agree. Changed.

vrozov · 2018-02-07T03:52:50Z

-        <exclusion>
-          <artifactId>calcite-avatica</artifactId>
-          <groupId>org.apache.calcite</groupId>
+          <groupId>org.json</groupId>


Has new version of hive introduced the dependency on org.json:json?

The master branch of Hive doesn't have org.json.
But Hive2.1 includes this:
https://github.com/apache/hive/blob/branch-2.1/pom.xml#L608
https://github.com/apache/hive/blob/branch-2.1/pom.xml#L608
and so on...

vrozov · 2018-02-07T03:56:55Z

+    <dependency>
+      <groupId>org.apache.parquet</groupId>
+      <artifactId>parquet-column</artifactId>
+      <version>${parquet.version}</version>


Any other parquet dependencies? If they are not needed, why the explicit dependency on org.apache.parquet:parquet-column is necessary?

Hive parquet dependencies cause an issues in Drill. So relocating is decision.
But in this case one Drill test DRILL-3938 is failed, because Parquet 1.8.0 version started to prohibit empty struct/groups on MessageType level PARQUET-278.
But from 1.8.1 version it is allowed again - PARQUET-363.

So Drill parquet-column version without throwing an Exception solves the issue.
For Hive2.3 it can be changed.

Why is it safe to change one library? Will it be safer to upgrade all hive dependencies on parquet to the same version?

hive-exec uses only two parquet dependencies: parquet-column and parquet-hadoop-bundle.
But Drill doesn't use own version of parquet-hadoop-bundle and moreover Drill version for it is absent in maven repository.

It appears that Hive 2.3.2 uses parquet-column 1.8.1 version as well. But last Apache Hive master is updated to 1.9.0 version.
I have added comment about it into drill-hive-exec-shaded POM to update it in future.

vrozov · 2018-02-07T03:58:10Z

        <groupId>org.apache.maven.plugins</groupId>
        <artifactId>maven-shade-plugin</artifactId>
-        <version>2.1</version>
+        <version>3.1.0</version>


What is the reason for the version change and should it be applied to other modules where shading is used?

It is a last stable version of this plugin. Update is not important.
We use 2.4 version around the project. But here was an older version.

Can it be unified in the plugin management of the Drill root pom?

Good idea. Done

vdiravka

@vrozov I have addressed your comments. Please take a look.

vdiravka · 2018-02-07T15:22:09Z

          <groupId>commons-codec</groupId>
          <artifactId>commons-codec</artifactId>
        </exclusion>
+        <exclusion>


version conflict.

vdiravka · 2018-02-07T15:24:02Z

    <dependency>
      <groupId>org.apache.calcite</groupId>
      <artifactId>calcite-core</artifactId>
+      <version>${calcite.version}</version>


hive-exec needs own calcite 1.6 version.
Calcite version in DependencyManagement of root pom leads to using Drill Calcite version over the whole project.

vdiravka · 2018-02-07T15:24:15Z

    <dependency>
      <groupId>org.apache.calcite</groupId>
      <artifactId>calcite-core</artifactId>
+      <version>${calcite.version}</version>


vdiravka · 2018-02-07T15:51:12Z

+   * @param properties table or partition properties
+   * @param sd storage descriptor
+   */
+  public static void verifyAndAddTransactionalProperties(JobConf job, Properties properties, StorageDescriptor sd) {


job involves table properties, but JobConf hasn't any method to return that properties. It is possible to get only one property. But looks like this is not an issue.
There is in AcidUtils the isTablePropertyTransactional method with Configuration input parameter.
Changed. Thank you.

vdiravka · 2018-02-07T15:51:33Z

+
+      // Try to get get column names and types from table or partition properties. If they are absent there, get columns
+      // data from storage descriptor of the table
+      if (properties.containsKey(serdeConstants.LIST_COLUMNS) && properties.containsKey(serdeConstants.LIST_COLUMN_TYPES)) {


Agree. Changed.

vdiravka · 2018-02-07T16:18:06Z

+        colNames = job.get(serdeConstants.LIST_COLUMNS);
+        colTypes = job.get(serdeConstants.LIST_COLUMN_TYPES);
+      } else {
+        final StringBuilder colNamesBuilder = new StringBuilder();


Thank you, it's cleaner.

vdiravka · 2018-02-07T16:26:01Z

-        <exclusion>
-          <artifactId>calcite-avatica</artifactId>
-          <groupId>org.apache.calcite</groupId>
+          <groupId>org.json</groupId>


The master branch of Hive doesn't have org.json.
But Hive2.1 includes this:
https://github.com/apache/hive/blob/branch-2.1/pom.xml#L608
https://github.com/apache/hive/blob/branch-2.1/pom.xml#L608
and so on...

vdiravka · 2018-02-07T16:46:30Z

+    <dependency>
+      <groupId>org.apache.parquet</groupId>
+      <artifactId>parquet-column</artifactId>
+      <version>${parquet.version}</version>


Hive parquet dependencies cause an issues in Drill. So relocating is decision.
But in this case one Drill test DRILL-3938 is failed, because Parquet 1.8.0 version started to prohibit empty struct/groups on MessageType level PARQUET-278.
But from 1.8.1 version it is allowed again - PARQUET-363.

So Drill parquet-column version without throwing an Exception solves the issue.
For Hive2.3 it can be changed.

vdiravka · 2018-02-07T16:58:13Z

        <groupId>org.apache.maven.plugins</groupId>
        <artifactId>maven-shade-plugin</artifactId>
-        <version>2.1</version>
+        <version>3.1.0</version>


It is a last stable version of this plugin. Update is not important.
We use 2.4 version around the project. But here was an older version.

vrozov · 2018-02-08T05:57:24Z

+          colNamesList.add(col.getName());
+          colTypesList.add(col.getType());
+        }
+        colNames = Joiner.on(",").join(colNamesList);


Consider Joiner.on(",").join(Iterables.transform(sd.getCols(), toName)); where toName is

private static Function<FieldSchema, String> toName = new Function<FieldSchema, String>() { @Nullable @Override public String apply(@Nullable FieldSchema input) { return input.getName(); } };

I have changed it. But we need call input.getName() and input.getType(), that's why I use two Functions and code became bigger. Once we will use Java8 it can be smaller. Or did I miss something here?

vdiravka

@vrozov I have addressed your comments. Changes are in a new commit. This is a description of the changes:

Apache Hive version client updating to 2.3.2 version.
"io.dropwizard.metrics:metrics-core" with last 4.0.2 version is added to dependencyManagement block in Drill root POM
Exclusion of "hive-exec" in "hive-hbase-handler" is already in Drill root dependencyManagement POM
Hive Calcite libraries are excluded (Calcite CBO was disabled in firs commit)
"jackson-core" dependency is added to DependencyManagement block in Drill root POM file.
For Hive 2.1 client older "com.fasterxml.jackson.core:jackson-databind" is included
Exclusion of "log4j:log4j" dependency from "hive-exec" is replaced to dependencyManagement block of Drill root POM file.
Exclusion of "com.codahale.metrics:metrics-core" from "hive-hbase-handler" is removed.

One Advanced test started to mysteriously fail
framework/resources/Advanced/tpch/tpch_sf100/parquet/01.q for some decimal values, not for hive but for parquet. I will investigate the reason.

vdiravka · 2018-02-13T17:45:00Z

          <groupId>commons-codec</groupId>
          <artifactId>commons-codec</artifactId>
        </exclusion>
+        <exclusion>


Both cases can resolve it. In details io.dropwizard.metrics:metrics-core is nowhere used in Drill. And this is a dependency for tephra-core and transitive for hive-metastore. But it conflicts with Drill's com.codahale.metrics.
hive-metastore uses 3.0.1 version of this dependency, but the last version in maven repository is 4.0.2.
I added this dependency to dependencyManagement block with 4.0.2 version and conflict is resolved as well. I think it is a better decision, because it can help to avoid similar conflicts in future.

Also metrics-core in hive-hbase-handler has not influence to Drill, so I've removed my exclusion of it.

vdiravka · 2018-02-13T17:45:49Z

    <dependency>
      <groupId>org.apache.calcite</groupId>
      <artifactId>calcite-core</artifactId>
+      <version>${calcite.version}</version>


I have returned calcite-core version into DependencyManagement block. Drill Calcite version libraries are included into "drill-hive-exec-shaded" module.

I will tell you why it works for now:
When user submits query in Drill via Hive plugin, the query is validated and planned via Drill Calcite. So Hive Calcite isn't necessary for it.
Hive Calcite is used only in the process of Drill unit testing, where a lot of Hive specific queries are performed to setup Hive store for testing. But Drill Calcite and Avatica versions have conflicts with Hive old Calcite and Avatica versions. That's why I have disabled Calcite cost based optimizator conf.set(ConfVars.HIVE_CBO_ENABLED.varname, "false");. We can enable it again once Hive will leverage the newest Calcite version. The comment about it is added into drill-hive-exec-shaded POM.

vdiravka · 2018-02-13T17:49:58Z

+          colNamesList.add(col.getName());
+          colTypesList.add(col.getType());
+        }
+        colNames = Joiner.on(",").join(colNamesList);


I have changed it. But we need call input.getName() and input.getType(), that's why I use two Functions and code became bigger. Once we will use Java8 it can be smaller. Or did I miss something here?

vdiravka · 2018-02-13T17:53:09Z

+    <dependency>
+      <groupId>org.apache.parquet</groupId>
+      <artifactId>parquet-column</artifactId>
+      <version>${parquet.version}</version>


hive-exec uses only two parquet dependencies: parquet-column and parquet-hadoop-bundle.
But Drill doesn't use own version of parquet-hadoop-bundle and moreover Drill version for it is absent in maven repository.

It appears that Hive 2.3.2 uses parquet-column 1.8.1 version as well. But last Apache Hive master is updated to 1.9.0 version.
I have added comment about it into drill-hive-exec-shaded POM to update it in future.

vrozov · 2018-02-20T15:10:24Z


  // SerDe of the reading partition (or table if the table is non-partitioned)
-  protected SerDe partitionSerDe;
+  protected Deserializer partitionSerDe;


rename to 'partitionDeserializer'

vrozov · 2018-02-20T15:19:29Z

      HiveUtilities.addConfToJob(job, partitionProperties);

-      final SerDe tableSerDe = createSerDe(job, table.getSd().getSerdeInfo().getSerializationLib(), tableProperties);
+      final Deserializer tableSerDe = createSerDe(job, table.getSd().getSerdeInfo().getSerializationLib(), tableProperties);


rename both function and the variable.

Done. tableSerDe -> tableDeserializer and createSerDe -> createDeserializer
If you meant getSerdeInfo() as well, that function is a part of hive code.

vrozov · 2018-02-20T15:28:47Z

        </exclusion>
      </exclusions>
    </dependency>
+    <!--Once newer hive-exec version leverages parquet-column 1.9.0, this dependency can be deleted -->


Can it be moved to the dependency management if this is still necessary?

vrozov · 2018-02-20T15:43:37Z

        <groupId>org.apache.maven.plugins</groupId>
        <artifactId>maven-shade-plugin</artifactId>
-        <version>2.1</version>
+        <version>3.1.0</version>


Can it be unified in the plugin management of the Drill root pom?

vdiravka

@vrozov Comments have been addressed.
I have rebased the branch onto the master branch.
The above-mentioned Advanced test failure is test-framework issue.

vdiravka · 2018-02-20T17:30:43Z


  // SerDe of the reading partition (or table if the table is non-partitioned)
-  protected SerDe partitionSerDe;
+  protected Deserializer partitionSerDe;


vdiravka · 2018-02-20T17:30:49Z

      HiveUtilities.addConfToJob(job, partitionProperties);

-      final SerDe tableSerDe = createSerDe(job, table.getSd().getSerdeInfo().getSerializationLib(), tableProperties);
+      final Deserializer tableSerDe = createSerDe(job, table.getSd().getSerdeInfo().getSerializationLib(), tableProperties);


Done. tableSerDe -> tableDeserializer and createSerDe -> createDeserializer
If you meant getSerdeInfo() as well, that function is a part of hive code.

vdiravka · 2018-02-20T17:31:18Z

        </exclusion>
      </exclusions>
    </dependency>
+    <!--Once newer hive-exec version leverages parquet-column 1.9.0, this dependency can be deleted -->


vdiravka · 2018-02-20T17:31:43Z

        <groupId>org.apache.maven.plugins</groupId>
        <artifactId>maven-shade-plugin</artifactId>
-        <version>2.1</version>
+        <version>3.1.0</version>


Good idea. Done

vrozov

LGTM, please squash.

vdiravka · 2018-02-21T11:26:40Z

@vrozov Thank you for CR.
Commits are squashed, Jira and PR are renamed to 2.3.2 version upgrade.

…2.1.2-mapr-1710 versions respectively * Improvements to allow of reading Hive bucketed transactional ORC tables; * Updating hive properties for tests and resolving dependencies and API conflicts: - Fix for "hive.metastore.schema.verification", MetaException(message: Version information not found in metastore) https://cwiki.apache.org/confluence/display/Hive/Hive+Schema+Tool METASTORE_SCHEMA_VERIFICATION="false" property is added - Added METASTORE_AUTO_CREATE_ALL="true", properties to tests, because some additional tables are necessary in Hive metastore - Disabling calcite CBO for (Hive's CalcitePlanner) for tests, because it is in conflict with Drill's Calcite version for Drill unit tests. HIVE_CBO_ENABLED="false" property - jackson and parquet libraries are relocated in hive-exec-shade module - org.apache.parquet:parquet-column Drill version is added to "hive-exec" to allow of using Parquet empty group on MessageType level (PARQUET-278) - Removing of commons-codec exclusion from hive core. This dependency is necessary for hive-exec and hive-metastore. - Setting Hive internal properties for transactional scan: HiveConf.HIVE_TRANSACTIONAL_TABLE_SCAN and for schema evolution: HiveConf.HIVE_SCHEMA_EVOLUTION, IOConstants.SCHEMA_EVOLUTION_COLUMNS, IOConstants.SCHEMA_EVOLUTION_COLUMNS_TYPES - "io.dropwizard.metrics:metrics-core" with last 4.0.2 version is added to dependencyManagement block in Drill root POM - Exclusion of "hive-exec" in "hive-hbase-handler" is already in Drill root dependencyManagement POM - Hive Calcite libraries are excluded (Calcite CBO was disabled) - "jackson-core" dependency is added to DependencyManagement block in Drill root POM file - For MapR Hive 2.1 client older "com.fasterxml.jackson.core:jackson-databind" is included - "log4j:log4j" dependency is excluded from "hive-exec", "hive-metastore", "hive-hbase-handler".

amansinha100 · 2018-02-24T16:39:53Z

+1. merged the PR in commit 27aa236.

vdiravka force-pushed the DRILL-5978 branch from 476c44c to b0317c9 Compare February 5, 2018 21:16

vdiravka changed the title ~~Upgrade drill-hive libraries to 2.1.1 version.~~ DRILL-5978: Upgrade drill-hive libraries to 2.1.1 version. Feb 5, 2018

vdiravka changed the title ~~DRILL-5978: Upgrade drill-hive libraries to 2.1.1 version.~~ DRILL-5978: Upgrade Hive libraries to 2.1.1 version. Feb 5, 2018

sohami reviewed Feb 6, 2018

View reviewed changes

vdiravka commented Feb 6, 2018

View reviewed changes

vdiravka force-pushed the DRILL-5978 branch from 8d8c331 to d70a7fb Compare February 6, 2018 14:31

vrozov reviewed Feb 7, 2018

View reviewed changes

vdiravka commented Feb 7, 2018

View reviewed changes

vdiravka force-pushed the DRILL-5978 branch from 30c6243 to c403136 Compare February 7, 2018 17:11

vrozov reviewed Feb 8, 2018

View reviewed changes

vdiravka force-pushed the DRILL-5978 branch 2 times, most recently from ab7130d to 96cd7a2 Compare February 16, 2018 16:32

vdiravka commented Feb 16, 2018

View reviewed changes

vrozov reviewed Feb 20, 2018

View reviewed changes

vdiravka force-pushed the DRILL-5978 branch 2 times, most recently from b798969 to 9ab2b76 Compare February 20, 2018 22:58

vdiravka commented Feb 20, 2018

View reviewed changes

vrozov reviewed Feb 20, 2018

View reviewed changes

vdiravka force-pushed the DRILL-5978 branch 2 times, most recently from 53e75ce to 8834db7 Compare February 21, 2018 11:16

vdiravka changed the title ~~DRILL-5978: Upgrade Hive libraries to 2.1.1 version.~~ DRILL-5978: Upgrade Hive libraries to 2.3.2 version Feb 21, 2018

vdiravka changed the title ~~DRILL-5978: Upgrade Hive libraries to 2.3.2 version~~ DRILL-5978: Updating of Apache and MapR Hive libraries to 2.3.2 and 2.1.2-mapr-1710 versions respectively Feb 21, 2018

vdiravka force-pushed the DRILL-5978 branch from 8834db7 to 571e6dc Compare February 21, 2018 20:49

asfgit closed this in 27aa236 Feb 24, 2018

asfimport mentioned this pull request Jun 23, 2024

enforce non empty group on MessageType level apache/parquet-java#1807

Closed

Uh oh!

Conversation

vdiravka commented Feb 5, 2018

Uh oh!

priteshm commented Feb 5, 2018 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

vdiravka left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

vdiravka Feb 13, 2018 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

priteshm commented Feb 5, 2018 •

edited

Loading

vdiravka Feb 13, 2018 •

edited

Loading