Documentation – Redise Pack

A guide to Redise Pack installation, operation and administration

open all | close all open all | close all

Overview

What is Redise Pack (RP)?

Redise Pack (RP) is based on the proven technology behind the Redise offering, used by thousands of customers.

You can install RP in your environment of choice, whether in an on-premises data-center or your preferred cloud platform, giving you full control of your data and configuration.

Note: The previous name of Redise Pack (RP) was “Redis Labs Enterprise Cluster” or RLEC for short. So you may see this on our web site or other places.

What does RP enable?

RP’s architecture supports multiple databases created by multiple users across the same cluster infrastructure while keeping the databases completely isolated from one another.

In addition, RP gives you the flexibility to run your Redis database in multiple configurations to accommodate your specific performance and availability needs.

Each such Redis database can separately conform to one of the following deployment templates:

  • Standalone (a single master Redis server)
  • Highly-available (a single master with a single slave)
  • Cluster (multiple master shards)
  • Highly-available cluster (multiple master and slave shards)

Redise Flash

RP is available on Flash with the same blazing fast sub-millisecond latencies attained when running Redis on RAM, while guaranteeing 100% compatibility with all Redis clients, data types, and commands, at just a fraction of the cost.

The underlying technology of Redise Flash uses Flash as a RAM extension, rather than as persistent storage. The global key list and all ‘hot’ values are kept in RAM, while all ‘cold’ values (those that typically account for the larger part of the dataset) are kept in Flash.

Redise Flash is extremely suitable for scenarios in which values are significantly larger than the keys, and where 10-40% of all objects are “hot” objects.

In many cases, Redise Flash can save up to 70% in resource costs as compared to an all-RAM deployment.

For additional details, refer to Redise Flash.

Features

Seamless scalability

With RP, a dataset can grow beyond the largest node in the cluster and be processed by any number of cores. By creating a Redis cluster and sharding your dataset across multiple nodes (using a sharding policy of choice), RP overcomes the memory limitations of a single node and the performance limitations of a single core. Dynamic scaling is easy; you can increase your dataset’s maximum size by simply changing the memory limit setting, or increase the number of shards with the click of a button. Scaling, upgrades and downgrades do not incur any downtime.

True high availability

In addition to replicating a database within the same data center, you can also replicate a dataset across data-centers and across regions, to achieve high availability, disaster recovery and performance benefits. If a node fails, the data is seamlessly served from a replacement node in the cluster without human intervention. RP technology is capable of automatically handling node failures, Redis failures and proxy failures. Auto-failover is carried out within a few seconds.

Built-in data persistence, backups and replication

RP enables the use of Redis AOF (Append-Only File) every second or every write, or snapshots of your dataset every 1, 6 or 12 hours to persistent storage. In addition, you can back up your dataset periodically or ad-hoc to an FTP server or to AWS S3. Other cloud storage options, such as Azure Geo-Redundant Storage, SoftLayer Object Storage and Google Cloud Storage will be added in a future release.

Predictable high-performance

Datasets are processed by multiple cores to guarantee the best performance. In addition, RP uses advanced mechanisms to guarantee high-performance, even during intensive disk-access scenarios.

Multiple dedicated databases

You can run multiple databases over a single RP deployment, each running in a dedicated process and in a non-blocking manner.

Unlimited database connections

Use as many database connections as you want.

Full-featured management UI

RP provides a full-featured and easy-to-use management user interface (UI) for setting up, configuring and monitoring the cluster and each database. In addition, RP provides configurable alert notifications for important events at the single database level as well as the cluster level.

Cluster automation

RP’s fully-managed solution takes care of all your database scaling, data-persistence tuning, shards migration and auto-failover needs.

24/7 support

Paying customers enjoy our premium 24/7 support via Redis Labs’ online or phone helpdesk.