fix: handle dmap continuous execution in profiling framework

dmap is designed to run continuously and report MapReduce results at
intervals, which caused it to hang during profiling. Fixed by:

- Added run_profile_dmap() function that runs dmap in background
- Sends SIGINT after 3 seconds to cleanly exit dmap
- Updated all dmap profiling calls to use the new function
- Applied fix to both profile_benchmarks.sh and profile_dmap.sh

This ensures dmap can be profiled successfully without timing out.

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <noreply@anthropic.com>

2 files changed, 55 insertions, 10 deletions

diff --git a/benchmarks/profile_benchmarks.sh b/benchmarks/profile_benchmarks.sh
index b0bcf64..976aaf7 100755
--- a/benchmarks/profile_benchmarks.sh
+++ b/benchmarks/profile_benchmarks.sh
@@ -69,6 +69,38 @@ run_profile() {
     echo
 }
 
+# Special function for profiling dmap which runs continuously
+run_profile_dmap() {
+    local cmd=$1
+    local name=$2
+    local args=$3
+    
+    echo -e "${GREEN}Profiling $cmd - $name${NC}"
+    
+    for i in $(seq 1 $PROFILE_RUNS); do
+        echo "  Run $i/$PROFILE_RUNS..."
+        echo "  Command: $cmd -profile -profiledir $PROFILE_DIR $args (will interrupt after 3s)"
+        
+        # Run dmap in background, wait a bit for it to process, then interrupt it
+        $cmd -profile -profiledir "$PROFILE_DIR" $args > /dev/null 2>&1 &
+        local pid=$!
+        
+        # Wait for dmap to process the file and generate initial results
+        sleep 3
+        
+        # Send interrupt signal to make it exit cleanly
+        kill -INT $pid 2>/dev/null
+        wait $pid 2>/dev/null
+        
+        echo "  Completed"
+        
+        # Small delay between runs
+        sleep 1
+    done
+    
+    echo
+}
+
 # Generate test data
 echo -e "${GREEN}Preparing test data...${NC}"
 generate_test_data "1MB" "$TEST_DATA_DIR/small.log"
@@ -132,13 +164,13 @@ if [ ! -f "$TEST_DATA_DIR/dtail_format.log" ]; then
 fi
 
 # Profile dmap with DTail format  
-run_profile "../dmap" "simple_count" "-plain -cfg none -query 'from STATS select count(*)' -files $TEST_DATA_DIR/dtail_format.log"
-run_profile "../dmap" "aggregations" "-plain -cfg none -query 'from STATS select sum(\$goroutines),avg(\$cgocalls),max(lifetimeConnections)' -files $TEST_DATA_DIR/dtail_format.log"
-run_profile "../dmap" "group_by_connections" "-plain -cfg none -query 'from STATS select currentConnections,count(*) group by currentConnections' -files $TEST_DATA_DIR/dtail_format.log"
+run_profile_dmap "../dmap" "simple_count" "-plain -cfg none -query 'from STATS select count(*)' -files $TEST_DATA_DIR/dtail_format.log"
+run_profile_dmap "../dmap" "aggregations" "-plain -cfg none -query 'from STATS select sum(\$goroutines),avg(\$cgocalls),max(lifetimeConnections)' -files $TEST_DATA_DIR/dtail_format.log"
+run_profile_dmap "../dmap" "group_by_connections" "-plain -cfg none -query 'from STATS select currentConnections,count(*) group by currentConnections' -files $TEST_DATA_DIR/dtail_format.log"
 
 # Also test CSV format
 echo -e "\n${YELLOW}Testing CSV format with dmap${NC}"
-run_profile "../dmap" "csv_query" "-plain -cfg none -query 'select user,action,count(*) where status=\"success\" group by user,action logformat csv' -files $TEST_DATA_DIR/test.csv"
+run_profile_dmap "../dmap" "csv_query" "-plain -cfg none -query 'select user,action,count(*) where status=\"success\" group by user,action logformat csv' -files $TEST_DATA_DIR/test.csv"
 
 echo
 echo -e "${GREEN}Profiling complete!${NC}"
diff --git a/benchmarks/profile_dmap.sh b/benchmarks/profile_dmap.sh
index a3a1151..904c793 100755
--- a/benchmarks/profile_dmap.sh
+++ b/benchmarks/profile_dmap.sh
@@ -100,20 +100,33 @@ echo -e "${GREEN}Profiling dmap queries...${NC}"
 # Query 1: Simple count
 echo -e "\n${YELLOW}Query: Count by hostname${NC}"
 QUERY="from STATS select count(\$line) group by hostname outfile $TEST_DATA_DIR/count_output.csv"
-echo "Command: timeout 10s ../dmap -profile -profiledir $PROFILE_DIR -plain -cfg none -query \"$QUERY\" -files $TEST_DATA_DIR/stats_small.log"
-timeout 10s ../dmap -profile -profiledir "$PROFILE_DIR" -plain -cfg none -query "$QUERY" -files "$TEST_DATA_DIR/stats_small.log" 2>&1 | head -10
+echo "Command: ../dmap -profile -profiledir $PROFILE_DIR -plain -cfg none -query \"$QUERY\" -files $TEST_DATA_DIR/stats_small.log (will interrupt after 3s)"
+# Run dmap in background and interrupt after 3 seconds
+../dmap -profile -profiledir "$PROFILE_DIR" -plain -cfg none -query "$QUERY" -files "$TEST_DATA_DIR/stats_small.log" 2>&1 | head -10 &
+DMAP_PID=$!
+sleep 3
+kill -INT $DMAP_PID 2>/dev/null
+wait $DMAP_PID 2>/dev/null
 
 # Query 2: Aggregations
 echo -e "\n${YELLOW}Query: Sum and average${NC}"
 QUERY="from STATS select sum(\$goroutines),avg(\$goroutines) group by hostname outfile $TEST_DATA_DIR/sum_avg_output.csv"
-echo "Command: timeout 10s ../dmap -profile -profiledir $PROFILE_DIR -plain -cfg none -query \"$QUERY\" -files $TEST_DATA_DIR/stats_small.log"
-timeout 10s ../dmap -profile -profiledir "$PROFILE_DIR" -plain -cfg none -query "$QUERY" -files "$TEST_DATA_DIR/stats_small.log" 2>&1 | head -10
+echo "Command: ../dmap -profile -profiledir $PROFILE_DIR -plain -cfg none -query \"$QUERY\" -files $TEST_DATA_DIR/stats_small.log (will interrupt after 3s)"
+../dmap -profile -profiledir "$PROFILE_DIR" -plain -cfg none -query "$QUERY" -files "$TEST_DATA_DIR/stats_small.log" 2>&1 | head -10 &
+DMAP_PID=$!
+sleep 3
+kill -INT $DMAP_PID 2>/dev/null
+wait $DMAP_PID 2>/dev/null
 
 # Query 3: Min/Max
 echo -e "\n${YELLOW}Query: Min and max${NC}"
 QUERY="from STATS select min(currentConnections),max(lifetimeConnections) group by hostname outfile $TEST_DATA_DIR/min_max_output.csv"
-echo "Command: timeout 10s ../dmap -profile -profiledir $PROFILE_DIR -plain -cfg none -query \"$QUERY\" -files $TEST_DATA_DIR/stats_small.log"
-timeout 10s ../dmap -profile -profiledir "$PROFILE_DIR" -plain -cfg none -query "$QUERY" -files "$TEST_DATA_DIR/stats_small.log" 2>&1 | head -10
+echo "Command: ../dmap -profile -profiledir $PROFILE_DIR -plain -cfg none -query \"$QUERY\" -files $TEST_DATA_DIR/stats_small.log (will interrupt after 3s)"
+../dmap -profile -profiledir "$PROFILE_DIR" -plain -cfg none -query "$QUERY" -files "$TEST_DATA_DIR/stats_small.log" 2>&1 | head -10 &
+DMAP_PID=$!
+sleep 3
+kill -INT $DMAP_PID 2>/dev/null
+wait $DMAP_PID 2>/dev/null
 
 echo
 echo -e "${GREEN}Analyzing dmap profiles...${NC}"