Configure Phoenix server scan page timeout

trinodb · Jan 15, 2025 · 1b1359c · 1b1359c
1 parent 13e51e1
commit 1b1359c
Show file tree

Hide file tree

Showing 5 changed files with 39 additions and 6 deletions.
diff --git a/docs/src/main/sphinx/connector/phoenix.md b/docs/src/main/sphinx/connector/phoenix.md
@@ -45,11 +45,12 @@ to load custom Phoenix client connection properties.
 
 The following Phoenix-specific configuration properties are available:
 
-| Property name                 | Required | Description                                                                                                                                                                                                                                                                                                                                                                                      |
-| ----------------------------- | -------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ |
-| `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.) |
+| Property name                      | Required | Description                                                                                                                                                                                                                                                                                                                                                                                      |
+|------------------------------------| -------- |--------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
+| `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.server-scan-page-timeout` | No       | The time limit on the amount of work single RPC request can do before it times out. Type: [](prop-type-duration).                                                                                                                                                                                                                                                                                |
 
 ```{include} jdbc-common-configurations.fragment
 ```

diff --git a/plugin/trino-phoenix5/pom.xml b/plugin/trino-phoenix5/pom.xml
@@ -53,6 +53,11 @@
             <artifactId>log</artifactId>
         </dependency>
 
+        <dependency>
+            <groupId>io.airlift</groupId>
+            <artifactId>units</artifactId>
+        </dependency>
+
         <dependency>
             <groupId>io.trino</groupId>
             <artifactId>trino-base-jdbc</artifactId>

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
@@ -90,6 +90,7 @@
 import static io.trino.plugin.phoenix5.PhoenixClient.DEFAULT_DOMAIN_COMPACTION_THRESHOLD;
 import static io.trino.plugin.phoenix5.PhoenixErrorCode.PHOENIX_CONFIG_ERROR;
 import static java.util.Objects.requireNonNull;
+import static org.apache.phoenix.query.QueryServices.PHOENIX_SERVER_PAGE_SIZE_MS;
 import static org.apache.phoenix.util.ReadOnlyProps.EMPTY_PROPS;
 import static org.weakref.jmx.guice.ExportBinder.newExporter;
 
@@ -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);
+        config.getServerScanPageTimeout().ifPresent(duration -> connectionProperties.setProperty(PHOENIX_SERVER_PAGE_SIZE_MS, String.valueOf(duration.toMillis())));
         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
@@ -17,11 +17,13 @@
 import io.airlift.configuration.Config;
 import io.airlift.configuration.ConfigDescription;
 import io.airlift.configuration.validation.FileExists;
+import io.airlift.units.Duration;
 import jakarta.validation.constraints.Max;
 import jakarta.validation.constraints.Min;
 import jakarta.validation.constraints.NotNull;
 
 import java.util.List;
+import java.util.Optional;
 
 public class PhoenixConfig
 {
@@ -41,6 +43,11 @@ 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 Optional<Duration> serverScanPageTimeout = Optional.empty();
+
     @NotNull
     public String getConnectionUrl()
     {
@@ -94,4 +101,17 @@ public PhoenixConfig setReuseConnection(boolean reuseConnection)
         this.reuseConnection = reuseConnection;
         return this;
     }
+
+    public Optional<Duration> getServerScanPageTimeout()
+    {
+        return serverScanPageTimeout;
+    }
+
+    @Config("phoenix.server-scan-page-timeout")
+    @ConfigDescription("Phoenix scan page timeout to reflect the time limit on the amount of work single Scan RPC request can do")
+    public PhoenixConfig setServerScanPageTimeout(Duration serverScanPageTimeout)
+    {
+        this.serverScanPageTimeout = Optional.ofNullable(serverScanPageTimeout);
+        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
@@ -15,6 +15,7 @@
 
 import com.google.common.collect.ImmutableList;
 import com.google.common.collect.ImmutableMap;
+import io.airlift.units.Duration;
 import org.junit.jupiter.api.Test;
 
 import java.io.IOException;
@@ -25,6 +26,7 @@
 import static io.airlift.configuration.testing.ConfigAssertions.assertFullMapping;
 import static io.airlift.configuration.testing.ConfigAssertions.assertRecordedDefaults;
 import static io.airlift.configuration.testing.ConfigAssertions.recordDefaults;
+import static java.util.concurrent.TimeUnit.HOURS;
 
 public class TestPhoenixConfig
 {
@@ -35,7 +37,8 @@ public void testDefaults()
                 .setConnectionUrl(null)
                 .setResourceConfigFiles(ImmutableList.of())
                 .setMaxScansPerSplit(20)
-                .setReuseConnection(true));
+                .setReuseConnection(true)
+                .setServerScanPageTimeout(null));
     }
 
     @Test
@@ -49,12 +52,14 @@ public void testExplicitPropertyMappings()
                 .put("phoenix.config.resources", configFile.toString())
                 .put("phoenix.max-scans-per-split", "1")
                 .put("query.reuse-connection", "false")
+                .put("phoenix.server-scan-page-timeout", "11h")
                 .buildOrThrow();
 
         PhoenixConfig expected = new PhoenixConfig()
                 .setConnectionUrl("jdbc:phoenix:localhost:2181:/hbase")
                 .setResourceConfigFiles(ImmutableList.of(configFile.toString()))
                 .setMaxScansPerSplit(1)
+                .setServerScanPageTimeout(new Duration(11, HOURS))
                 .setReuseConnection(false);
 
         assertFullMapping(properties, expected);