Introduction
============

Navigating the analytics accelerator
------------------------------------

The accelerator organizes content into four areas:

- **Conceptual foundations** Build your understanding of analytics
  principles and EDB’s approach.

- **EDB core analytics technologies** Learn about EDB solutions and
  technologies that power our analytics offerings.

- **Practical guidance and solutions** Find use cases, persona-based
  guides, how-to articles, and tutorials.

- **Product-specific implementations** Access documentation for how
  these analytics capabilities surface and are managed in EDB products,
  such as EDB Hybrid Manager.

Conceptual foundations
----------------------

Understand the principles and strategies behind modern data analytics
and EDB’s approach.

- :ref:`Generic analytics concepts <Analytics Accelerator generic concepts>` 

Learn about data architectures (Data Warehouse, Data Lake, Lakehouse)
and foundational technologies (columnar storage, vectorized engines, and
others).

- :ref:`EDB analytics concepts <Analytics Accelerator concepts>` 

Explore EDB’s vision for Postgres® analytics and how EDB leverages core
technologies.

- :ref:`Explained: Analytics <learn/explained>` 

Review in-depth explanations of EDB analytical features, design choices,
and advanced topics. *(Coming soon)*

EDB core analytics technologies
-------------------------------

Learn about EDB’s analytics technologies and how they extend Postgres®.

- :ref:`EDB Postgres Lakehouse <EDB Postgres Lakehouse>` 

Review the EDB Postgres® Lakehouse solution and its components for
enabling analytics on object storage.

- :ref:`Apache Iceberg <Apache Iceberg>` 

Understand how EDB solutions use Apache Iceberg to manage large
analytical datasets.

- :ref:`Delta Lake <Delta Lake>` 

Learn how EDB Postgres® interacts with Delta tables to enable reliable
data lakes.

- :ref:`EDB Postgres Distributed (PGD) and Tiered Tables <EDB Postgres Distributed (PGD) and Tiered Tables>` 

Manage data across storage tiers using EDB Postgres Distributed (PGD)
and Lakehouse capabilities to optimize cost and performance.

Practical guidance and solutions
--------------------------------

Apply EDB’s analytics capabilities to meet your needs.

- :ref:`Analytics Accelerator for your role: a persona-based guide <a persona-based guide>` 

Follow learning paths for DBAs, DevOps engineers, data scientists, and
application developers.

Product-specific implementations
--------------------------------

Review how EDB analytics concepts and technologies are implemented in
EDB products.

- :ref:`Introduction <Introduction>` 

Access documentation for analytics features in EDB Hybrid Manager. This
includes HM Lakehouse clusters, using Iceberg, Delta, and tiered tables
in HM, and HM-specific tutorials.

Where to start
--------------

- Start with :ref:`Generic analytics concepts <Analytics Accelerator generic concepts>`  and :ref:`EDB Postgres Lakehouse <EDB Postgres Lakehouse>`  to understand core ideas.

- Explore practical guidance when available.

- Use product-specific documentation when working with EDB Hybrid
  Manager.

Postgres Lakehouse is built using a number of technologies:

- PostgreSQL

- `Seafowl <https://seafowl.io/>`_  , an analytical database 

- `Apache DataFusion <https://datafusion.apache.org/>`_  , the query engine used by Seafowl 

- `Delta Lake <https://delta.io>`_  (and specifically  `delta-rs <https://github.com/delta-io/delta-rs>`_  ), for implementing the storage and retrieval layer of Delta Tables 

Level 100
^^^^^^^^^

The most important thing to understand about Postgres Lakehouse is that
it separates storage from compute. This design allows you to scale them
independently, which is ideal for analytical workloads where queries can
be unpredictable and spiky. You wouldn’t want to keep a machine mostly
idle just to hold data on its attached hard drives. Instead, you can
keep data in object storage (and also in highly compressible formats),
and only provision the compute needed to query it when necessary.

|Level 100 Architecture|

On the compute side, a vectorized query engine is optimized to query
Lakehouse tables but still fall back to Postgres for full compatibility.

On the storage side, Lakehouse tables are stored using highly
compressible columnar storage formats optimized for analytics.

Level 200
^^^^^^^^^

Here’s a slightly more comprehensive diagram of how these services fit
together:

|Level 200 Architecture|

Level 300
^^^^^^^^^

Here’s the more detailed, zoomed-in view of “what’s in the box”:

|Level 300 Architecture|

.. |Level 100 Architecture| image:: /images/level-100.png
   :width: 70% 
   :target: images/level-100.png
.. |Level 200 Architecture| image:: /images/level-200.png
   :width: 70% 
   :target: images/level-200.png
.. |Level 300 Architecture| image:: /images/level-300.png
   :width: 70% 
   :target: images/level-300.png