use-cases/compare/hazelcast.html - ignite-website

<!--

▄▄▄ ██▓███ ▄▄▄ ▄████▄ ██░ ██ ▓█████ ██▓ ▄████ ███▄ █ ██▓▄▄▄█████▓▓█████

▒████▄ ▓██░ ██▒▒████▄ ▒██▀ ▀█ ▓██░ ██▒▓█ ▀ ▓██▒ ██▒ ▀█▒ ██ ▀█ █ ▓██▒▓ ██▒ ▓▒▓█ ▀

▒██ ▀█▄ ▓██░ ██▓▒▒██ ▀█▄ ▒▓█ ▄ ▒██▀▀██░▒███ ▒██▒▒██░▄▄▄░▓██ ▀█ ██▒▒██▒▒ ▓██░ ▒░▒███

░██▄▄▄▄██ ▒██▄█▓▒ ▒░██▄▄▄▄██ ▒▓▓▄ ▄██▒░▓█ ░██ ▒▓█ ▄ ░██░░▓█ ██▓▓██▒ ▐▌██▒░██░░ ▓██▓ ░ ▒▓█ ▄

▓█ ▓██▒▒██▒ ░ ░ ▓█ ▓██▒▒ ▓███▀ ░░▓█▒░██▓░▒████▒ ░██░░▒▓███▀▒▒██░ ▓██░░██░ ▒██▒ ░ ░▒████▒

▒▒ ▓▒█░▒▓▒░ ░ ░ ▒▒ ▓▒█░░ ░▒ ▒ ░ ▒ ░░▒░▒░░ ▒░ ░ ░▓ ░▒ ▒ ░ ▒░ ▒ ▒ ░▓ ▒ ░░ ░░ ▒░ ░

▒ ▒▒ ░░▒ ░ ▒ ▒▒ ░ ░ ▒ ▒ ░▒░ ░ ░ ░ ░ ▒ ░ ░ ░ ░ ░░ ░ ▒░ ▒ ░ ░ ░ ░ ░

░ ▒ ░░ ░ ▒ ░ ░ ░░ ░ ░ ▒ ░░ ░ ░ ░ ░ ░ ▒ ░ ░ ░

░ ░ ░ ░░ ░ ░ ░ ░ ░ ░ ░ ░ ░ ░ ░ ░

-->

<!--

Licensed to the Apache Software Foundation (ASF) under one

or more contributor license agreements. See the NOTICE file

distributed with this work for additional information

regarding copyright ownership. The ASF licenses this file

to you 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.

-->

<!DOCTYPE html>

<html>

<head>

<title>Apache Ignite Data Grid vs Hazelcast Data Grid</title>

</head>

<body>

<h2 class="first">Ignite vs Hazelcast Data Grid</h2>

Both, Apache Ignite and Hazelcast provide a feature rich data grid functionality responsible for

partitioning and caching data in memory with the capability to scale out across distributed clusters.

<a href="/benchmarks/ignite-vs-hazelcast.html">Click Here for Benchmark Results </a>

</div>

</section>

<h3>Feature Comparison</h3>

Apache Ignite and Hazelcast have many differences in the way caching, transactions, and data querying

are supported. Below we outline only major differences, which in our opinion are most important whenever

choosing an in-memory data grid product.

<thead>

<tr>

<th style="text-align: center;font-size: larger">Apache Ignite</th>

<th style="text-align: center;font-size: larger">Hazelcast</th>

</tr>

</thead>

<tbody>

<tr><td colspan="2" style="text-align: center">Open Source Focus</td></tr>

<tr>

Ignite has been growing and enhancing its open source feature set. The latest

major additions include Ignite Fabric for C++ and .NET/C#, with

Node.JS integration coming soon as well. We have also added extremely fast

deadlock-free transactions.

</td>

Most of the development in Hazelcast is focused on their enterprise (paid)

offering. Features, like off-heap memory, or

continuous queries are only available in the paid edition,

with SSL-encryption feature completely removed from the open

source to the paid edition.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">JCache (JSR 107)</td></tr>

<tr>

Ignite is fully compliant with JCache (JSR 107) specification for caching.

</td>

Hazelcast is fully compliant with JCache (JSR 107) specification for caching.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Off-Heap Memory</td></tr>

<tr>

Ignite supports storing data in on-heap or off-heap memory,

depending on user configuration.

</td>

Hazelcast supports off-heap memory only in its commercial (paid) offering.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Off-Heap Indexes</td></tr>

<tr>

Whenever configured with off-heap memory, Ignite will store query indexes

off-heap as well (in order not to affect on-heap memory used by user applications).

</td>

Hazelcast does not have support for off-heap indexes.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Continuous Queries</td></tr>

<tr>

Ignite provides support for fail-safe continuous queries, i.e. allows clients and servers

to subscribe for continuous notifications to the data updates.

</td>

Hazelcast offers support for continuous queries only in its commercial (paid) edition.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">SSL Encryption</td></tr>

<tr>

Ignite provides support for SSL/TLS encryption, i.e. allows all clients and servers

to communicate over encrypted SLL or TLS protocols.

</td>

Hazelcast removed SLL encryption support from open source edition and now

offers it only in its commercial (paid) edition.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">SQL Queries</td></tr>

<tr>

Ignite supports complete SQL (ANSI-99) syntax for querying in-memory data.

</td>

Hazelcast has very limited support for SQL (only a few keywords).

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Queries with JOINs</td></tr>

<tr>

Ignite SQL includes full support for SQL JOINs, including

JOINs across multiple caches, for example:

select * from A a, B b where a.b_id = b.id

</td>

Hazelcast does not have any support for JOIN queries (with or without SQL).

Whenever needed, users perform JOINs by manually combining multiple query results.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Query Consistency</td></tr>

<tr>

Ignite provides full query consistency, i.e. queries are executed on a certain snapshot

of data, and updates that happened after the query execution started do not affect

the query result.

</td>

Hazelcast queries are inconsistent and it is possible that part of the query result will

see a certain update and another part will not.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Query Fault Tolerance</td></tr>

<tr>

Ignite queries are fault-tolerant, i.e. query result is always consistent and is not

affected by cluster topology changes, like nodes joining, leaving, or crashing.

</td>

Hazelcast queries are not fault-tolerant, i.e. query results become inconsistent

during cluster topology changes, while the data is being re-balanced

in the background.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Data Consistency</td></tr>

<tr>

Ignite supports atomic and transactional consistency on all in-memory

data, regardless of whether data is stored in PARTITIONED or REPLICATED

caches.

</td>

Hazelcast supports atomic and transactional consistency only on

PARTITIONED caches. Data stored in REPLICATED caches does not have

any transactional guarantees at all.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Deadlock-Free Transactions</td></tr>

<tr>

Ignite supports deadlock-free, optimistic transactions, which do not

acquire any locks, and free users from worrying about the lock order.

Such transactions also provide much better performance.

</td>

With Hazelcast, you always need to worry about updating data in the same order to

avoid deadlocks, which is often impossible, especially in large projects.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Transactional EntryProcessor</td></tr>

<tr>

Ignite supports transactional EntryProcessor, which allows to execute

collocated user logic on the server side, within a transaction.

</td>

With Hazelcast, it is impossible to use EntryProcessor or execute

user logic within a transaction.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Web Session Clustering</td></tr>

<tr>

Ignite supports caching and clustering web-sessions for all known application servers.

</td>

Hazelcast offers web-session clustering only in its commercial (paid) edition.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Compute Grid</td></tr>

<tr>

Ignite provides support for map-reduce, fork-join,

and basic distributed lambda processing on the cluster, including job load-balancing,

fault-tolerance, checkpointing, scheduling, etc.

</td>

Hazelcast only provides support for map-reduce and random job distribution within a cluster.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Streaming Grid</td></tr>

<tr>

Ignite provides support for in-memory streaming, including support for maintaining and

querying sliding windows of streaming data.

</td>

Hazelcast does not offer any support for streaming.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Service Grid</td></tr>

<tr>

Ignite allows users to easily cluster-enable their services, including support for various

cluster-singletons.

</td>

Hazelcast managed services do not provide support for cluster-singletons.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Fabric for .NET/C#</td></tr>

<tr>

Ignite provides full in-memory fabric APIs for .NET/C# users, including

ability to execute C# closures, C# caching, transaction, and querying APIs,

native C# CacheStore APIs, etc.

</td>

Hazelcast provides .NET/C# limited client APIs only.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Fabric for C++</td></tr>

<tr>

Ignite provides full in-memory fabric APIs for C++ users.

</td>

Hazelcast provides C++ limited client APIs only.

</td>

</tr>

<tr><td colspan="2" style="text-align: center">Fabric for Node.JS</td></tr>

<tr>

Ignite provides client API support for Node.JS users (coming in 1.5).

</td>

Hazelcast does not have any support for Node.JS.

</td>

</tr>

</tbody>

</table>

</section>

</main>

</div>

</body>

</html>