Skip to content

fix: mark Lance HMS tables as EXTERNAL to prevent data purge on DROP #142

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
xingsuo-zbz wants to merge 1 commit into
base: main
Choose a base branch
from
Open

fix: mark Lance HMS tables as EXTERNAL to prevent data purge on DROP #142

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions java/lance-namespace-hive2/src/main/java/org/lance/namespace/hive2/Hive2Util.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -124,6 +124,8 @@ public static Map<String, String> createLanceTableParams(Map<String, String> pro
}
params.put("table_type", "lance");
params.put("managed_by", "storage");
// HMS's MetaStoreUtils.isExternalTable() reads parameters.EXTERNAL, not Table.tableType.
params.put("EXTERNAL", "TRUE");
return params;
}
}
55 changes: 55 additions & 0 deletions java/lance-namespace-hive2/src/test/java/org/lance/namespace/hive2/TestHive2Namespace.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -16,6 +16,7 @@
import org.lance.namespace.LanceNamespace;
import org.lance.namespace.errors.LanceNamespaceException;
import org.lance.namespace.model.CreateNamespaceRequest;
import org.lance.namespace.model.DeclareTableRequest;
import org.lance.namespace.model.DescribeNamespaceRequest;
import org.lance.namespace.model.DescribeNamespaceResponse;
import org.lance.namespace.model.DescribeTableRequest;
Expand All @@ -30,6 +31,7 @@
import org.apache.arrow.memory.BufferAllocator;
import org.apache.arrow.memory.RootAllocator;
import org.apache.hadoop.hive.conf.HiveConf;
import org.apache.hadoop.hive.metastore.api.Table;
import org.assertj.core.util.Lists;
import org.junit.jupiter.api.AfterAll;
import org.junit.jupiter.api.AfterEach;
Expand Down Expand Up @@ -302,4 +304,57 @@ public void testCloseBeforeInitializeIsSafe() {
Hive2Namespace uninitialized = new Hive2Namespace();
uninitialized.close();
}

@Test
public void testDeclareTableMarksTableAsExternal() throws Exception {
CreateNamespaceRequest nsRequest = new CreateNamespaceRequest();
nsRequest.setId(Lists.list("ext_db"));
nsRequest.setMode("Create");
namespace.createNamespace(nsRequest);

DeclareTableRequest declareRequest = new DeclareTableRequest();
declareRequest.setId(Lists.list("ext_db", "ext_tbl"));
declareRequest.setLocation(tmpDirBase + "/ext_db/ext_tbl");
namespace.declareTable(declareRequest);

Table hmsTable = metastore.clientPool().run(client -> client.getTable("ext_db", "ext_tbl"));
assertEquals("EXTERNAL_TABLE", hmsTable.getTableType());
assertEquals("TRUE", hmsTable.getParameters().get("EXTERNAL"));
assertEquals("lance", hmsTable.getParameters().get("table_type"));
assertEquals("storage", hmsTable.getParameters().get("managed_by"));
}

/**
* Verifies that a table declared via {@link org.lance.namespace.LanceNamespace#declareTable}
* survives a raw HMS {@code dropTable(deleteData=true)} — i.e. the same call path used by Hive
* CLI / Spark {@code DROP TABLE}. Without {@code parameters.EXTERNAL=TRUE}, HMS would treat the
* table as MANAGED and purge the data directory.
*/
@Test
public void testDropTablePreservesDataForDeclaredTable() throws Exception {
CreateNamespaceRequest nsRequest = new CreateNamespaceRequest();
nsRequest.setId(Lists.list("keep_db"));
nsRequest.setMode("Create");
namespace.createNamespace(nsRequest);

File location = new File(tmpDirBase, "keep_db/keep_tbl");
assertTrue(location.mkdirs());
File dataFile = new File(location, "data.marker");
assertTrue(dataFile.createNewFile());

DeclareTableRequest declareRequest = new DeclareTableRequest();
declareRequest.setId(Lists.list("keep_db", "keep_tbl"));
declareRequest.setLocation(location.getAbsolutePath());
namespace.declareTable(declareRequest);

metastore
.clientPool()
.run(
client -> {
client.dropTable("keep_db", "keep_tbl", true, true);
return null;
});

assertTrue(dataFile.exists(), "data file was purged: EXTERNAL=TRUE not honored by HMS");
}
}
2 changes: 2 additions & 0 deletions java/lance-namespace-hive3/src/main/java/org/lance/namespace/hive3/Hive3Util.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -175,6 +175,8 @@ public static Map<String, String> createLanceTableParams(Map<String, String> pro
}
params.put("table_type", "lance");
params.put("managed_by", "storage");
// HMS's MetaStoreUtils.isExternalTable() reads parameters.EXTERNAL, not Table.tableType.
params.put("EXTERNAL", "TRUE");
return params;
}
}
46 changes: 46 additions & 0 deletions java/lance-namespace-hive3/src/test/java/org/lance/namespace/hive3/TestHive3Namespace.java
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -17,6 +17,7 @@
import org.lance.namespace.errors.InvalidInputException;
import org.lance.namespace.errors.LanceNamespaceException;
import org.lance.namespace.model.CreateNamespaceRequest;
import org.lance.namespace.model.DeclareTableRequest;
import org.lance.namespace.model.DescribeNamespaceRequest;
import org.lance.namespace.model.DescribeNamespaceResponse;
import org.lance.namespace.model.DescribeTableRequest;
Expand All @@ -31,6 +32,7 @@
import org.apache.arrow.memory.BufferAllocator;
import org.apache.arrow.memory.RootAllocator;
import org.apache.hadoop.hive.conf.HiveConf;
import org.apache.hadoop.hive.metastore.api.Table;
import org.assertj.core.util.Lists;
import org.junit.jupiter.api.AfterAll;
import org.junit.jupiter.api.AfterEach;
Expand Down Expand Up @@ -416,4 +418,48 @@ public void testCloseBeforeInitializeIsSafe() {
Hive3Namespace uninitialized = new Hive3Namespace();
uninitialized.close();
}

@Test
public void testDeclareTableMarksTableAsExternal() throws Exception {
DeclareTableRequest declareRequest = new DeclareTableRequest();
declareRequest.setId(Lists.list("test_catalog", "test_db", "ext_tbl"));
declareRequest.setLocation(tmpDirBase + "/test_catalog/test_db/ext_tbl");
namespace.declareTable(declareRequest);

Table hmsTable =
metastore.clientPool().run(client -> client.getTable("test_catalog", "test_db", "ext_tbl"));
assertEquals("EXTERNAL_TABLE", hmsTable.getTableType());
assertEquals("TRUE", hmsTable.getParameters().get("EXTERNAL"));
assertEquals("lance", hmsTable.getParameters().get("table_type"));
assertEquals("storage", hmsTable.getParameters().get("managed_by"));
}

/**
* Verifies that a table declared via {@link org.lance.namespace.LanceNamespace#declareTable}
* survives a raw HMS {@code dropTable(deleteData=true)} — i.e. the same call path used by Hive
* CLI / Spark {@code DROP TABLE}. Without {@code parameters.EXTERNAL=TRUE}, HMS would treat the
* table as MANAGED and purge the data directory.
*/
@Test
public void testDropTablePreservesDataForDeclaredTable() throws Exception {
File location = new File(tmpDirBase, "test_catalog/test_db/keep_tbl");
assertTrue(location.mkdirs());
File dataFile = new File(location, "data.marker");
assertTrue(dataFile.createNewFile());

DeclareTableRequest declareRequest = new DeclareTableRequest();
declareRequest.setId(Lists.list("test_catalog", "test_db", "keep_tbl"));
declareRequest.setLocation(location.getAbsolutePath());
namespace.declareTable(declareRequest);

metastore
.clientPool()
.run(
client -> {
client.dropTable("test_catalog", "test_db", "keep_tbl", true, true);
return null;
});

assertTrue(dataFile.exists(), "data file was purged: EXTERNAL=TRUE not honored by HMS");
}
}
Loading