Add config for the threshold of column count for auto analyze.

apache · Nov 28, 2023 · 8ae679d · 8ae679d
1 parent b48c40e
commit 8ae679d
Show file tree

Hide file tree

Showing 4 changed files with 26 additions and 0 deletions.
diff --git a/fe/fe-core/src/main/java/org/apache/doris/qe/SessionVariable.java b/fe/fe-core/src/main/java/org/apache/doris/qe/SessionVariable.java
@@ -434,6 +434,8 @@ public class SessionVariable implements Serializable, Writable {
 
     public static final String ENABLE_AUTO_ANALYZE = "enable_auto_analyze";
 
+    public static final String AUTO_ANALYZE_TABLE_WIDTH_THRESHOLD = "auto_analyze_table_width_threshold";
+
     public static final String FASTER_FLOAT_CONVERT = "faster_float_convert";
 
     public static final String ENABLE_DECIMAL256 = "enable_decimal256";
@@ -1315,6 +1317,13 @@ public void setEnableLeftZigZag(boolean enableLeftZigZag) {
             flag = VariableMgr.GLOBAL)
     public boolean enableAutoAnalyze = true;
 
+    @VariableMgr.VarAttr(name = AUTO_ANALYZE_TABLE_WIDTH_THRESHOLD,
+            description = {"参与自动收集的最大表宽度，列数多于这个参数的表不参与自动收集",
+                "Maximum table width to enable auto analyze, "
+                    + "table with more columns than this value will not be auto analyzed."},
+            flag = VariableMgr.GLOBAL)
+    public int autoAnalyzeTableWidthThreshold = 70;
+
     @VariableMgr.VarAttr(name = AUTO_ANALYZE_START_TIME, needForward = true, checker = "checkAnalyzeTimeFormat",
             description = {"该参数定义自动ANALYZE例程的开始时间",
                     "This parameter defines the start time for the automatic ANALYZE routine."},

diff --git a/fe/fe-core/src/main/java/org/apache/doris/statistics/StatisticConstants.java b/fe/fe-core/src/main/java/org/apache/doris/statistics/StatisticConstants.java
@@ -98,6 +98,8 @@ public class StatisticConstants {
 
     public static final int SUBMIT_JOB_LIMIT = 5;
 
+    public static final int AUTO_ANALYZE_TABLE_WIDTH_THRESHOLD = 70;
+
     static {
         SYSTEM_DBS.add(SystemInfoService.DEFAULT_CLUSTER
                 + ClusterNamespace.CLUSTER_DELIMITER + FeConstants.INTERNAL_DB_NAME);

diff --git a/fe/fe-core/src/main/java/org/apache/doris/statistics/StatisticsAutoCollector.java b/fe/fe-core/src/main/java/org/apache/doris/statistics/StatisticsAutoCollector.java
@@ -178,6 +178,11 @@ protected AnalysisInfo getReAnalyzeRequiredPart(AnalysisInfo jobInfo) {
         AnalysisManager analysisManager = Env.getServingEnv().getAnalysisManager();
         TableStatsMeta tblStats = analysisManager.findTableStatsStatus(table.getId());
 
+        // Skip tables that are too width.
+        if (table.getColumns().size() > StatisticsUtil.getAutoAnalyzeTableWidthThreshold()) {
+            return null;
+        }
+
         if (!table.needReAnalyzeTable(tblStats)) {
             return null;
         }

diff --git a/fe/fe-core/src/main/java/org/apache/doris/statistics/util/StatisticsUtil.java b/fe/fe-core/src/main/java/org/apache/doris/statistics/util/StatisticsUtil.java
@@ -936,4 +936,14 @@ public static int getAnalyzeTimeout() {
         return StatisticConstants.ANALYZE_TIMEOUT_IN_SEC;
     }
 
+    public static int getAutoAnalyzeTableWidthThreshold() {
+        try {
+            return findConfigFromGlobalSessionVar(SessionVariable.AUTO_ANALYZE_TABLE_WIDTH_THRESHOLD)
+                .autoAnalyzeTableWidthThreshold;
+        } catch (Exception e) {
+            LOG.warn("Failed to get value of auto_analyze_table_width_threshold, return default", e);
+        }
+        return StatisticConstants.AUTO_ANALYZE_TABLE_WIDTH_THRESHOLD;
+    }
+
 }