Thoth Storages

This library provides a library called thoth-storages used in project Thoth. The library exposes core queries and methods for PostgreSQL database as well as adapters for manipulating with Ceph via its S3 compatible API.

Installation and Usage

The library can be installed via pip or Pipenv from PyPI:

pipenv install thoth-storages

The library does not provide any CLI, it is rather a low level library supporting other parts of Thoth.

You can run prepared testsuite via the following command:

pipenv install --dev
pipenv run python3 setup.py test

# To generate docs:
pipenv run python3 setup.py build_sphinx

Running PostgreSQL locally

You can use docker-compose present in this repository to run a local PostgreSQL instance:

$ docker-compose up

After running the command above (make sure your big fat daemon is up using systemctl start docker), you should be able to access a local PostgreSQL instance at localhost:5432. This is also the default configuration for PostgreSQL’s adapter - you don’t need to provide GRAPH_SERVICE_HOST explicitly. The default configuration uses database named thoth which can be accessed using postgres user and postgres password (SSL is disabled).

The provided docker-compose has also PGweb enabled for to have an UI for the database content. To access it visit http://localhost:8081/.

The provided docker-compose does not use any volume. After you containers restart, the content will not be available anymore.

If you would like to experiment with PostgreSQL programatically, you can use the following code snippet as a starting point:

from thoth.storages import GraphDatabase

graph = GraphDatabase()
graph.connect()
# To clear database:
# graph.drop_all()
# To initialize schema in the graph database:
# graph.initialize_schema()

Schema adjustment in deployment

TBD.

Generate schema images

You can use shipped CLI thoth-storages to automatically generate schema images out of the current models:

# First, make sure you have dev packages installed:
pipenv install --dev
PYTHONPATH=. pipenv run python3 ./thoth-storages generate-schema

The command above will produce 2 images named schema.png and schema_cache.png. The first PNG file shows schema for the main PostgreSQL instance and the latter one, as the name suggests, shows how cache schema looks like.

Creating own performance indicators

You can create your own performance indicators. To create own performance indicator, create a script which tests desired functionality of a library. An example can be matrix multiplication script present in performance repository. This script can be supplied to Dependency Monkey to validate certain combination of libraries in desired runtime and buildtime environment or directly on Amun API which will run the given script using desired software and hardware configuration. Please follow instructions on how to create a performance script shown in the README of performance repo.

To create relevant models, adjust thoth/storages/graph/models_performance.py file and add your model. Describe parameters (reported in @parameters section of performance indicator result) and result (reported in @result). The name of class should match name which is reported by performance indicator run.

class PiMatmul(Base, BaseExtension, PerformanceIndicatorBase):
    """A class for representing a matrix multiplication micro-performance test."""

    # Device used during performance indicator run - CPU/GPU/TPU/...
    device = Column(String(128), nullable=False)
    matrix_size = Column(Integer, nullable=False)
    dtype = Column(String(128), nullable=False)
    reps = Column(Integer, nullable=False)
    elapsed = Column(Float, nullable=False)
    rate = Column(Float, nullable=False)

All the models use SQLAchemy. See docs for more info.

Online debugging of queries

You can print to logger all the queries that are performed to a PostgreSQL instance. To do so, set the following environment variable:

export THOTH_STORAGES_DEBUG_QUERIES=1

Online debugging of queries

You can print information about PostgreSQL adapter together with statisics on the graph cache and memory cache usage to logger (it has to have at least level INFO set). To do so, set the following environment variable:

export THOTH_STORAGES_LOG_STATS=1

These statistics will be printed once the database adapter is destructed.

Graph database cache

The implementation of this library also provides a cache to speed up queries to graph database. This cache is especially suitable for prod systems not to query for popular packages multiple times.

The cache can be created with shipped CLI tool:

# When using version from this Git repository:
PYTHONPATH=. THOTH_STORAGES_GRAPH_CACHE="cache.sqlite3" pipenv run ./thoth-storages graph-cache -c ../adviser/cache_conf.yaml

# When using a version installed from PyPI:
THOTH_STORAGES_GRAPH_CACHE="cache.sqlite3" thoth-storages graph-cache -c ../adviser/cache_conf.yaml

The command above creates a SQLite3 database which carries some of the data loaded from the PostgreSQL database which help resolver resolve software stacks faster. The path to cache can be supplied using environment variable THOTH_STORAGES_GRAPH_CACHE. By default, the module will create an in-memory SQLite3 database and will not persist it onto disk. If the configuration points to non-existing file, an SQLite3 database will be created and persisted onto disk with data which were added into it based on runtime usage. This naturally re-uses graph cache multiple times across runs (filled with the data needed) as expected.

Take a look at adviser repo, at cache_conf.yaml file specifically, to see how cache_conf.yaml file should be structured. An example could be:

python-packages:
 - thoth-storages
 - tensorflow

With the configuration above, the cache will be created. This cache will hold a serialized dependency graph of TensorFlow and thoth-storages packages, together with node information to effectively construct TensorFlow’s dependency graph for transitive queries.

Note only information which should not change over time are captured in the cache; for example, packages which were not yet resolved during cache creation are not added to cache so system explicitly asks for resolution results next time (they might be resolved meanwhile).

To enable inserts into graph cache, set THOTH_STORAGES_GRAPH_CACHE_INSERTS_DISABLED to 0 (the default value of 1 disables it). Disabling inserts might be benefitial in deployments where you want to avoid building cache (overhead needed to insert data into graph cache, checks of uniqueness of entries and cache index creation which in sum are expensive operations).

To disable graph cache completely, set THOTH_STORAGES_GRAPH_CACHE_DISABLED environment variable to 1 (the default value of 0 enables it).

Graph Database Schema

Graph database schema

Graph database schema is automatically generated by Dia. You can install it from repositories. Issue the following command in Fedora:

dnf install dia

You can find schema.dia in Git repository of thoth-storages repository. Do not forget to export changes made to schema diagram to the PNG file in the same directory.