Skip to content
Making Postgres and Elasticsearch work together like it's 2019
Branch: master-es5.6
Clone or download
Pull request Compare This branch is 264 commits ahead, 114 commits behind master.
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Type Name Latest commit message Commit time
Failed to load latest commit information.

ZomboDB Build Status

(This branch is for Elasticsearch 5.6.4)

Upgrading from ES 1.7 or ES 2.4?

In short, ZomboDB doesn't provide a migration process from ES 1.7/2.4 to ES 5.6.4. You'll need to drop all of your USING zombodb indexes, manually delete all the indexes from your ES cluster, upgrade your ES cluster to 5.6.4, upgrade the ZomboDB Postgres extension to v5.6.4-x.x.x and then re-create all your USING zombodb indexes.

If you require assistance with this process, please don't hesitate to contact ZomboDB, LLC.

What's Different with ES 5.6.4+ Support?

In general, everything is significantly faster.

Elasticsearch 5.6.4 is generally faster for indexing and searching, and its API has improved such that ZomboDB is able to implement certain optimizations that make everything faster.

All SQL-level syntax is backwards-compatible along with ZomboDB's query language syntax.

About ZomboDB

ZomboDB is a Postgres extension that enables efficient full-text searching via the use of indexes
backed by Elasticsearch. In order to achieve this, ZomboDB implements Postgres' Access Method API.

In practical terms, a ZomboDB index appears to Postgres as no different than a standard btree index. As such, standard SQL commands are fully supported, including SELECT, BEGIN, COMMIT, ABORT, INSERT, UPDATE, DELETE, COPY, and VACUUM.

Because ZomboDB implements Postgres' Access Method API, ZomboDB indexes are MVCC-safe, even as concurrent sessions mutate underlying tables. Following Postgres' MVCC rules means that every transaction sees, at all times, a view of the backing Elasticsearch index that is always consistent with the current transaction's snapshot.

Behind the scenes, ZomboDB indexes communicate with Elasticsearch via HTTP and are automatically synchronized, in a high-performance manner, as data changes.

Index management happens using standard Postgres SQL commands such as CREATE INDEX, REINDEX, and ALTER INDEX. Searching uses standard SQL SELECT statements with a custom operator that exposes a full-text query language supporting most of Elasticsearch's query constructs.

Elasticsearch-calculated aggregations are also provided through custom functions.

Quick Links


  • transaction-safe, MVCC-correct full text queries and Elasticsearch aggregations
  • managed & queried via standard Postgres SQL
  • works with tables of any structure
  • automatically creates Elasticsearch Mappings supporting most datatypes, including arrays
  • works with all Postgres query plans, including sequential scans
  • use whatever method you currently use for talking to Postgres (JDBC, DBI, libpq, etc)
  • extremely fast indexing
  • store document source in Elasticsearch so ZomboDB-generated indexes can by used by 3rd-party tools like Kibana
  • per-row scoring with term/phrase boosting
  • record count estimation
  • custom full-text query language supporting nearly all of Elasticsearch's search features, such as
    • boolean operations
    • proximity (in and out of order)
    • phrases, wildcards, fuzzy terms/phrases
    • regular expressions, inline scripts
    • range queries
    • "more like this"
    • any Elasticsearch query construct through direct JSON
  • query expansion, index linking, and block routing for improved cross-index join performance
  • search multiple tables at once
  • high-performance hit highlighting
  • limit/offset/sorting of fulltext queries by Elasticsearch
  • support for common Elasticsearch aggregations, including ability to nest
  • access to all of Elasticsearch's aggregations via direct JSON
  • extensive test suite

Not to suggest that these things are impossible, but there's a small set of non-features too:

  • ZomboDB indexes are not WAL-logged by Postgres. As such, ZomboDB indexes are not recoverable in the event of a Postgres server crash and will require a REINDEX
  • interoperability with various Postgres replication schemes is unknown
  • Postgres HOT update chains are not supported (necessitates a VACUUM FULL if a HOT-updated row is found during index creation)


Please visit to download.

If you want to integrate with some kind of CI or deployment system, you can intuit the pattern for versions from the Elasticsearch plugin and Postgres extension download links, but it'll be something like:

For the Postgres extension binaries, you'll need to use the one that's for your Postgres + Linux distro combination -- the example above is for Postgres 9.5 on Ubuntu Trusty.

What you need

Product Version
Postgres 9.3, 9.4, 9.5
Elasticsearch 5.6.4

For information about how to develop/build ZomboDB, see the Development Guide.

How to Use It

Usage is really quite simple. Note that this is just a brief overview. See the various documentation files for more detailed information.

Install the extension:


Create a table:

CREATE TABLE products (
    name text NOT NULL,
    keywords varchar(64)[],
    short_summary phrase,
    long_description fulltext, 
    price bigint,
    inventory_count integer,
    discontinued boolean default false,
    availability_date date
-- insert some data

Index it:

CREATE INDEX idx_zdb_products 
          ON products 
       USING zombodb(zdb('products', products.ctid), zdb(products))
        WITH (url='http://localhost:9200/', shards=5, replicas=1);

Query it:

  FROM products 
 WHERE zdb('products', ctid) ==> 'keywords:(sports,box) or long_description:(wooden w/5 away) and price < 100000';

Contact Information


The name is an homage to and its long history of continuous self-affirmation.

Development began in 2013 by Technology Concepts & Design, Inc as a closed-source effort to provide transaction safe full-text searching on top of Postgres tables. While Postgres' "tsearch" features are useful, they're not necessarily adequate for 200 column-wide tables with 100M rows, each containing large text content.

Initially designed on-top of Postgres' Foreign Data Wrapper API, ZomboDB quickly evolved into an index type (Access Method) so that queries are MVCC-safe and standard SQL can be used to query and manage indexes.

Elasticsearch was chosen as the backing search index because of its horizontal scaling abilities, performance, and general ease of use.

ZomboDB was open-sourced in July 2015 and has been used in numerous production systems of various sizes and complexity.

Credit and Thanks

Credit goes to Technology Concepts & Design, Inc, its management, and its development and quality assurance teams not only for their work during the early development days but also for their on-going support now that ZomboDB is open-source.


Portions Copyright 2013-2015 Technology Concepts & Design, Inc.
Portions Copyright 2015-2019 ZomboDB, LLC

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.

You can’t perform that action at this time.