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.
- Latest Release
- Installation instructions
- Getting Started Tutorial
- Index Management, Index Options, and Type Mapping
- Query Syntax
- SQL-level API
- v3.1.0 Release Notes
- transaction-safe, MVCC-correct full text queries
- managed & queried via standard Postgres SQL
- works with tables of any structure
- automatically creates Elasticsearch Mappings supporting most datatypes, including arrays
- supports full set of Elasticsearch language analyzers
- support for custom analyzer chains
- custom per-field mappings
- json/jsonb columns as nested objects for flexible schemaless sub-documents
- 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
- 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 and index linking
- support for SIREn to resolve index links
- 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)
Product | Version |
---|---|
Postgres | 9.3, 9.4, 9.5 |
Elasticsearch | 1.7.1+ (not 2.0) |
For information about how to develop/build ZomboDB, see the Development Guide.
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 EXTENSION zombodb;
Create a table:
CREATE TABLE products (
id SERIAL8 NOT NULL PRIMARY KEY,
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:
SELECT *
FROM products
WHERE zdb('products', ctid) ==> 'keywords:(sports,box) or long_description:(wooden w/5 away) and price < 100000';
- www.zombodb.com
- Google Group: [email protected]
- Twitter: @zombodb
- via github Issues and Pull Requests ;)
The name is an homage to zombo.com 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.
Two years later, it's been used in production systems for quite some time and is now open-source.
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 2015-2017 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
http://www.apache.org/licenses/LICENSE-2.0
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.