nodetool resetfullquerylog

Resets the full query logging path to the configured value.

---

Synopsis

nodetool [connection_options] resetfullquerylog

See connection options for connection options.

Description

nodetool resetfullquerylog stops full query logging and cleans up log files in the configured log directory (and any path specified via JMX). This command effectively disables FQL if it was enabled.

Cleans Log Directory

This command stops FQL and cleans log files in the configured directories. It does not simply reset the path—it deactivates logging and removes log data.

---

Examples

Basic Usage

nodetool resetfullquerylog

Reset and Verify

nodetool resetfullquerylog
nodetool getfullquerylog

---

Behavior

When resetfullquerylog is executed:

1. FQL is disabled if it was enabled
2. Log files in the configured directory are cleaned up
3. Log files in any JMX-specified path are also cleaned
4. The FQL configuration is reset to defaults

What Changes

Aspect	Before	After
FQL enabled	Yes (if was enabled)	No
Log files	Present	Cleaned
Log path	Custom or configured	Reset to config

---

When to Use

Restore Default Path

After using a temporary custom path:

# Previously enabled with custom path
nodetool enablefullquerylog --path /tmp/fql_debug

# Reset to configured path
nodetool resetfullquerylog

# Verify
nodetool getfullquerylog

Configuration Consistency

Align runtime settings with configuration file:

# Check current runtime path
nodetool getfullquerylog | grep log_dir

# Check configured path
grep "log_dir" /etc/cassandra/cassandra.yaml

# Reset to match config
nodetool resetfullquerylog

After Debugging Session

Return to normal operation after debug session:

# Debug session used custom path
# Now restore normal path
nodetool resetfullquerylog

# Continue with normal FQL path
nodetool getfullquerylog

---

Workflow: Debug Session with Custom Path

fql_debug_session.sh

#!/bin/bash
DEBUG_PATH="/tmp/fql_debug_$(date +%Y%m%d_%H%M%S)"
mkdir -p $DEBUG_PATH

echo "=== FQL Debug Session ==="

# 1. Save current configuration
echo "1. Current FQL configuration:"
nodetool getfullquerylog

# 2. Enable with debug path
echo ""
echo "2. Enabling FQL with debug path: $DEBUG_PATH"
nodetool enablefullquerylog --path $DEBUG_PATH --roll-cycle MINUTELY

# 3. Wait for data collection
echo ""
echo "3. Collecting data... Press Enter when done."
read

# 4. Disable FQL
echo ""
echo "4. Disabling FQL..."
nodetool disablefullquerylog

# 5. Process debug logs
echo ""
echo "5. Debug logs captured at: $DEBUG_PATH"
echo "   Size: $(du -sh $DEBUG_PATH)"
echo "   Entries: $(fqltool dump $DEBUG_PATH 2>/dev/null | wc -l)"

# 6. Reset to default path
echo ""
echo "6. Resetting to default path..."
nodetool resetfullquerylog

# 7. Verify reset
echo ""
echo "7. Configuration after reset:"
nodetool getfullquerylog

echo ""
echo "=== Debug Session Complete ==="
echo "Debug logs remain at: $DEBUG_PATH"

---

Configuration Reference

cassandra.yaml Setting

cassandra.yaml

full_query_logging_options:
    log_dir: /var/log/cassandra/fql
    # ... other settings

Runtime Override

# Override configured path at runtime
nodetool enablefullquerylog --path /custom/path

# Reset to cassandra.yaml path
nodetool resetfullquerylog

---

Cluster-Wide Operations

Reset on All Nodes

reset_fql_cluster.sh

#!/bin/bash
echo "Resetting FQL cluster-wide..."

# Get list of node IPs from local nodetool status
nodes=$(nodetool status | grep "^UN" | awk '{print $2}')

for node in $nodes; do
    echo -n "$node: "
    ssh "$node" 'nodetool resetfullquerylog 2>/dev/null && echo "reset" || echo "FAILED"'
done

echo ""
echo "Verification:"
for node in $nodes; do
    echo "=== $node ==="
    ssh "$node" 'nodetool getfullquerylog 2>/dev/null | grep "log_dir"'
done

---

Troubleshooting

Reset Has No Effect

# Check if cassandra.yaml has FQL configuration
grep -A 10 "full_query_logging_options" /etc/cassandra/cassandra.yaml

# If no config, reset has nothing to reset to
# The path may remain unchanged or empty

Path Not Changed After Reset

# Verify the reset command succeeded
nodetool resetfullquerylog

# Check current configuration
nodetool getfullquerylog

# Check cassandra.yaml for configured path
grep "log_dir" /etc/cassandra/cassandra.yaml

# If cassandra.yaml doesn't have a path configured,
# the runtime path may remain

Logs at Old Path

# After reset, existing logs remain at old path
# They are not moved or deleted

# Check old path
ls -la /old/custom/path/

# Move or archive if needed
mv /old/custom/path/* /archive/

---

Managing Multiple Paths

After Reset

cleanup_fql_paths.sh

#!/bin/bash
# Get current configured path
current_path=$(nodetool getfullquerylog | grep "log_dir" | awk '{print $2}')

echo "Current FQL path: $current_path"
echo ""

# Find other potential FQL directories
echo "Potential FQL directories on this system:"
find /var/log/cassandra /tmp -type d -name "*fql*" 2>/dev/null

echo ""
echo "Review and clean up old FQL directories as needed."

---

Best Practices

Reset Guidelines

1. Document custom paths - Track when and why custom paths are used
2. Clean up after reset - Archive or remove logs at old paths
3. Verify configuration - Check cassandra.yaml has valid FQL config
4. Cluster consistency - Reset on all nodes if custom path was cluster-wide
5. Check disk space - Ensure default path has sufficient space

When to Reset

• After temporary debugging sessions
• To align runtime with configuration
• After maintenance operations that used custom paths
• To restore normal operational logging

---

Related Commands

Command	Relationship
enablefullquerylog	Enable FQL (with optional custom path)
disablefullquerylog	Disable FQL
getfullquerylog	View FQL configuration