Phoenix Connector: Configure Page size while creating Connection

trinodb · Jan 12, 2025 · 2ea50e1 · 2ea50e1
1 parent 223828d
commit 2ea50e1
Show file tree

Hide file tree

Showing 4 changed files with 27 additions and 1 deletion.
diff --git a/docs/src/main/sphinx/connector/phoenix.md b/docs/src/main/sphinx/connector/phoenix.md
@@ -50,6 +50,7 @@ The following Phoenix-specific configuration properties are available:
 | `phoenix.connection-url`      | Yes      | `jdbc:phoenix[:zk_quorum][:zk_port][:zk_hbase_path]`. The `zk_quorum` is a comma separated list of ZooKeeper servers. The `zk_port` is the ZooKeeper port. The `zk_hbase_path` is the HBase root znode path, that is configurable using `hbase-site.xml`.  By default the location is `/hbase`                                                                                                   |
 | `phoenix.config.resources`    | No       | Comma-separated list of configuration files (e.g. `hbase-site.xml`) to use for connection properties.  These files must exist on the machines running Trino.                                                                                                                                                                                                                                     |
 | `phoenix.max-scans-per-split` | No       | Maximum number of HBase scans that will be performed in a single split. Default is 20. Lower values will lead to more splits in Trino. Can also be set via session propery `max_scans_per_split`. For details see: [https://phoenix.apache.org/update_statistics.html](https://phoenix.apache.org/update_statistics.html). (This setting has no effect when guideposts are disabled in Phoenix.) |
+| `phoenix.scan-page-size`      | No       | The time limit on the amount of work single RPC request can do before it times out.                                                                                                                                                                                                                                                                                                              |
 
 ```{include} jdbc-common-configurations.fragment
 ```

diff --git a/plugin/trino-phoenix5/src/main/java/io/trino/plugin/phoenix5/PhoenixClientModule.java b/plugin/trino-phoenix5/src/main/java/io/trino/plugin/phoenix5/PhoenixClientModule.java
@@ -71,6 +71,7 @@
 import org.apache.hadoop.fs.Path;
 import org.apache.phoenix.jdbc.ConnectionInfo;
 import org.apache.phoenix.jdbc.PhoenixDriver;
+import org.apache.phoenix.query.QueryServices;
 
 import java.sql.SQLException;
 import java.util.Map;
@@ -209,6 +210,7 @@ public static Properties getConnectionProperties(PhoenixConfig config)
 
         ConnectionInfo connectionInfo = ConnectionInfo.create(config.getConnectionUrl(), EMPTY_PROPS, new Properties());
         connectionInfo.asProps().asMap().forEach(connectionProperties::setProperty);
+        connectionProperties.setProperty(QueryServices.PHOENIX_SERVER_PAGE_SIZE_MS, String.valueOf(config.getPageSize()));
         return connectionProperties;
     }
 

diff --git a/plugin/trino-phoenix5/src/main/java/io/trino/plugin/phoenix5/PhoenixConfig.java b/plugin/trino-phoenix5/src/main/java/io/trino/plugin/phoenix5/PhoenixConfig.java
@@ -41,6 +41,11 @@ public class PhoenixConfig
     private int maxScansPerSplit = 20;
     private boolean reuseConnection = true;
 
+    /**
+     * By default, let Phoenix client derive Page size from HBase RPC timeout configs.
+     */
+    private int pageSize = -1;
+
     @NotNull
     public String getConnectionUrl()
     {
@@ -94,4 +99,19 @@ public PhoenixConfig setReuseConnection(boolean reuseConnection)
         this.reuseConnection = reuseConnection;
         return this;
     }
+
+    @Min(-1)
+    @Max(Integer.MAX_VALUE)
+    public int getPageSize()
+    {
+        return pageSize;
+    }
+
+    @Config("phoenix.scan-page-size")
+    @ConfigDescription("Phoenix Page size to reflect the time limit on the amount of work single RPC request can do")
+    public PhoenixConfig setPageSize(int pageSize)
+    {
+        this.pageSize = pageSize;
+        return this;
+    }
 }
diff --git a/plugin/trino-phoenix5/src/test/java/io/trino/plugin/phoenix5/TestPhoenixConfig.java b/plugin/trino-phoenix5/src/test/java/io/trino/plugin/phoenix5/TestPhoenixConfig.java
@@ -35,7 +35,8 @@ public void testDefaults()
                 .setConnectionUrl(null)
                 .setResourceConfigFiles(ImmutableList.of())
                 .setMaxScansPerSplit(20)
-                .setReuseConnection(true));
+                .setReuseConnection(true)
+                .setPageSize(-1));
     }
 
     @Test
@@ -49,12 +50,14 @@ public void testExplicitPropertyMappings()
                 .put("phoenix.config.resources", configFile.toString())
                 .put("phoenix.max-scans-per-split", "1")
                 .put("query.reuse-connection", "false")
+                .put("phoenix.scan-page-size", "11")
                 .buildOrThrow();
 
         PhoenixConfig expected = new PhoenixConfig()
                 .setConnectionUrl("jdbc:phoenix:localhost:2181:/hbase")
                 .setResourceConfigFiles(ImmutableList.of(configFile.toString()))
                 .setMaxScansPerSplit(1)
+                .setPageSize(11)
                 .setReuseConnection(false);
 
         assertFullMapping(properties, expected);