> For the complete documentation index, see [llms.txt](https://mariadb.com/docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://mariadb.com/docs/analytics/mariadb-columnstore/high-availability/optimization-and-tuning/profiling-memory-usage-with-heaptrack.md).

# Profiling Memory Usage with Heaptrack

Heaptrack is a memory profiler used to track allocation patterns, locate memory leaks, and analyze consumption spikes in MariaDB ColumnStore subprocesses, specifically `PrimProc`. The outputs help generate performance flame graphs for deep-dive engineering diagnostics.

## Installation

1. Clone the directory:

   ```bash
   git clone https://github.com/KDE/heaptrack.git
   ```
2. Install the development tools group:

   ```bash
   dnf group install "Development Tools" -y
   ```
3. Create the build folder layout:

   ```bash
   cd heaptrack
   mkdir build
   cd build
   ```
4. Install prerequisite library dependencies:

   ```bash
   yum install cmake libunwind libunwind-devel boost boost-devel -y
   ```
5. Configure the build files for a Release target:

   ```bash
   cmake -DCMAKE_BUILD_TYPE=Release .. # look for messages about missing dependencies!
   ```
6. Compile the tracking application:

   ```bash
   make -j$(nproc)
   ```
7. Verify the compiled tracking binary target path:

   ```bash
   ls -la /home/rocky/heaptrack/build/bin/heaptrack
   ```

## Running a Memory Profile

1. Stop the active process service:

   ```
   systemctl stop mcs-primproc
   ```
2. Relaunch the process in the foreground under Heaptrack:

   ```
   /home/rocky/heaptrack/build/bin/heaptrack /usr/bin/PrimProc -f
   ```

## Capturing and Sharing Results

1. Check system logs for process behaviors:

   ```bash
   journalctl -ru mcs-primproc
   ```
2. Stop the process service wrapper instance:

   ```bash
   systemctl stop mcs-primproc
   ```
3. Halt the tracing session: Press `Ctrl` + `C` in the active heaptrack window.
4. Share the compressed data profile with engineering: Locate and transfer the resulting archive artifact:

   ```
   /home/rocky/heaptrack/build/heaptrack.PrimProc.37747.zst
   ```

## Historical Engineering References

The following benchmarking and signal commands are preserved from internal engineering logs as testing options and should be adapted carefully based on your specific verification context:

{% code overflow="wrap" %}

```bash
# Simulating database workloads via sysbench:
# time sysbench /mnt/data/perf/s21.lua --mysql-db=tpc_h --mysql-user=root --skip_trx --events=8 --threads=5 --time=3600 run --histogram --report-interval=1

# Triggering a manual process crash context:
# kill -SIGHUP $(pidof PrimProc);
```

{% endcode %}


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://mariadb.com/docs/analytics/mariadb-columnstore/high-availability/optimization-and-tuning/profiling-memory-usage-with-heaptrack.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.