Metadata-Version: 2.1
Name: pyftdc
Version: 0.3.0
Summary: A small example package
Author: Jorge Imperial
Author-email: Jorge Imperial <jlimperial@protonmail.com>
Project-URL: Homepage, https://gitlab.com/jimper/mongo_ftdc
Project-URL: Bug Tracker, https://gitlab.com/jimper/issues
Classifier: Programming Language :: Python :: 3
Classifier: License :: OSI Approved :: MIT License
Classifier: Operating System :: OS Independent
Requires-Python: >=3.6
Description-Content-Type: text/markdown
Provides-Extra: test
License-File: LICENSE

pyftdc
==============

[![Gitter][gitter-badge]][gitter-link]

|      CI              | status |
|----------------------|--------|
| pip builds           | [![Pip Actions Status][actions-pip-badge]][actions-pip-link] |



A MongoDB FTDC files parser written in C++ that provides Python bindings using [pybind11](https://github.com/pybind/pybind11) and scikit-build.



[gitter-badge]:            https://badges.gitter.im/pybind/Lobby.svg
[gitter-link]:             https://gitter.im/pybind/Lobby
[actions-badge]:           https://github.com/pybind/pyftdc/workflows/Tests/badge.svg
[actions-conda-link]:      https://github.com/pybind/pyftdc/actions?query=workflow%3AConda
[actions-conda-badge]:     https://github.com/pybind/pyftdc/workflows/Conda/badge.svg
[actions-pip-link]:        https://github.com/pybind/pyftdc/actions?query=workflow%3APip
[actions-pip-badge]:       https://github.com/pybind/pyftdc/workflows/Pip/badge.svg
[actions-wheels-link]:     https://github.com/pybind/pyftdc/actions?query=workflow%3AWheels
[actions-wheels-badge]:    https://github.com/pybind/pyftdc/workflows/Wheels/badge.svg



Requisites
------------

To build the source distribution, you will need Python 3.8 or newer, git, python3-dev/python3-devel installed.

Please read the [build document](docs/build.md) for more details.


Installation
------------

**Building on Unix (Ubuntu, Centos, macOS)**

  
 1. clone this repository and change to the top level directory.
      ```
      git clone git@gitlab.com:jimper/mongo_ftdc.git 
      cd mongo_ftdc
      ```
      
 2. Install Python libraries to build binaries. Create a virtual environment to make your life easier.
 
      ```
      python3 -m venv venv
      source venv/bin/activate
      pip3 install --user-pep517 .
      ```
      
    You will now have built and installed in your virtual environment.
    

Alternatively, you can use setup.py directly, but for that you will need to manually install the required libraries into your virtual environment by running

     
     cd mongo_ftdc
     pip install -r requirements.txt
     
     
After which you can create a source distribution or a binary wheel:

     
     python3 setup.py sdist
     python3 setup.py bdist_wheel
     
These will reside in the _dist_ directory.


**Building on Windows**
  
  Not tested yet, but 'It should work'(TM)
  


License
-------

Apache V2

Test call
---------

```python
import pyftdc


def get_prefixed_metrics_names(param, ds):
    ops_counters = []
    for name in ds.metrics_names:
        if name.startswith(param):
            ops = ds.get_metric(name)
            ops_counters.append((name, ops))

    return ops_counters


def open_files_in_dir(dir_path, prefix):
    from os import listdir

    files_read = []
    try:
        dir_list = listdir(dir_path)
        for file_name in dir_list:
            if file_name.startswith(prefix):
                parser = pyftdc.FTDCParser()
                ds = parser.parse_file(dir_path + '/' + file_name)
                if ds:
                    print(f'File: {ds.file}')
                    print(f'{ds.metadata}')
                    ts = ds.get_metric("start")
                    if ts:
                        ts_size = len(ts)

                        print(f'Timestamp count {ts_size}. Start:{ts[0]}  Last: {ts[-1]}')

                        op_counter_names = get_prefixed_metrics_names('serverStatus.opcounters', ds)
                        cpu = get_prefixed_metrics_names('systemMetrics.cpu', ds)
                        disk = get_prefixed_metrics_names('systemMetrics.disks.nvme1n1', ds)

                        xxx = ds.get_metric_list_numpy(['systemMetrics.cpu.iowait_ms', 'xxx', 'systemMetrics.cpu.num_cpus'])
                        disk_n = ds.get_metric_numpy('systemMetrics.disks.nvme1n1.writes')

                        files_read.append(file_name)
                    else:
                        print(f'No timestamps on this dataset.')
    except FileNotFoundError as not_found:
        print('Path not found.')

    return files_read


def open_dir(dir_path):
    multi_parser = pyftdc.FTDCParser()
    datasets = multi_parser.parse_dir(dir_path)
    ds_count = len(datasets)
    print(f'There are {ds_count} datasets from {dir_path}')
    if ds_count > 0:
        for ds in datasets:
            if ds:
                print(f'File: {ds.file}')
                print(f'{ds.metadata}')
                ts = ds.get_metrics("start")
                if ts:
                    ts_size = len(ts)

                    print(f'Timestamp count {ts_size}. Start:{ts[0]}  Last: {ts[-1]}')

                    op_counter_names = get_prefixed_metrics_names('serverStatus.opcounters', ds)
                    cpu = get_prefixed_metrics_names('systemMetrics.cpu', ds)
                    disk = get_prefixed_metrics_names('systemMetrics.disks.nvme1n1', ds)

                    print('')
                else:
                    print(f'No timestamps on this dataset.')
            else:
                print(f'Bad dataset from file {ds}')


if __name__ == "__main__":
    files = open_files_in_dir('/somepath/diagnostic.data/',
                              'metrics.2022-11-13T21')
    print(files)

```

[`cibuildwheel`]:          https://cibuildwheel.readthedocs.io
