Production Readiness Guidelines¶

The goal of this document is to have a checklist that production customers can use to make sure their deployment adheres to ScyllaDB’s recommendations. Before deploying to production you should follow up on each of the main bullets described below to verify they comply with the recommendations provided. Click the links for more information on each topic.

Before You Begin ¶

Pre-Deployment Requirements ¶

Scylla System Requirements - verify your instances, system, OS, etc are supported by Scylla for production machines.
Scylla Getting Started

Choose a Compaction Strategy ¶

Each workload may require a specific strategy. Refer to Choose a Compaction Strategy for details. If you are using Scylla Enterprise, in all cases ICS is preferred over STCS.

Incremental Compaction Strategy (ICS)¶

If you are using Scylla Enterprise, we highly recommend using ICS (the default setting) for any table that you have, or considering to have STCS as a compaction strategy. You will have much less Space Amplification with ICS as it only requires 25% additional storage, as opposed to STCS which requires 50% more.

Note

ICS is the default compaction strategy setting for Scylla Enterprise versions 2020.1 and higher.

Refer to Incremental Compaction Strategy for an overview of the benefits.
Refer to Incremental Compaction Strategy Overview for a description of how it works.

Resiliency ¶

When rolling out to production it is important to make sure your data is recoverable and your database can function anytime there is a power or equipment failure.

Replication Factors ¶

Verify the Replication Factor is set properly for each keyspace.

We recommend using an RF of at least three.

If you have a multi-datacenter architecture we recommend to have RF=3 on each DC.

For additional information:

Read more about Scylla Fault Tolerance
Take a course at Scylla University on RF.

Consistency Levels ¶

Verify the Consistency Level (CL) is set properly for each table.

We recommend using LOCAL_QUORUM across the cluster and DCs

For additional information:

Refer to Scylla Fault Tolerance
Watch a Demo
Take a course at Scylla University on CL

Gossip Configuration ¶

Choose the correct Snitch.

Always use GossipingPropertyFileSnitch or Ec2MultiRegionSnitch Do Not use SimpleStrategy on any production machine, even if you only have a single DC.

For additional information:
- Refer to Gossip in Scylla
- Follow the How to Switch Snitches procedure if required.
- Take a course at Scylla University on Gossip
Use the correct Data Replication strategy.

Use NetworkTopologyStrategy replication-strategy as it supports multi-DC for your keyspaces.

Seed Node Redundancy ¶

Each DC needs to have multiple seed nodes.

We recommend having 2-3 seed nodes per DC.

For additional information:

Refer to Scylla Seed Nodes
In the future, Scylla will be Seedless

Performance ¶

Verify you have run scylla_setup in order to tune ScyllaDB to your hardware.

If you are running on a physical hardware please take a look into the following configuration files:

perftune.yaml
cpuset.conf

perftune.yaml ¶

If you have more than 8 cores or 16 vcpu always use mode: sq_split

cpuset.conf ¶

Make sure that the configuration in /etc/scylla.d/cpuset.conf corresponds to sq_split and that the hyperthreads of physical core 0 are excluded from the CPU list.

Compression ¶

Inter-node Compression ¶

Enable Inter-node Compression by editing the Scylla Configuration file (/etc/scylla.yaml).

internode_compression: all

For additional information, see the Admin Guide Inter-node Compression section.

Driver Compression ¶

This refers to compressing traffic between the client and Scylla. Verify your client driver is using compressed traffic when connected to Scylla. As compression is driver settings dependent, please check your client driver manual or Scylla Drivers.

Connectivity ¶

Drivers Settings ¶

Use shard aware drivers wherever possible. Scylla Drivers (not third-party drivers) are shard aware.
Configure connection pool - open more connections (>3 per shard) and/Or more clients. See this blog.

Management ¶

You must use both Scylla Manager and Scylla Monitor.

Scylla Manager ¶

Scylla Manager enables centralized cluster administration and database automation such as repair, backup, and health-check.

Repair ¶

Run repairs preferably once a week and run them exclusively from Scylla Manager. Refer to Repair a Cluster

Backup and Restore ¶

We recommend the following:

Run a full weekly backup from Scylla Manager
Run a daily backup from Scylla Manager
Check the bucket used for restore. This can be done by performing a restore and making sure the data is valid. This action should be done once a month, or more frequently if needed. Ask our Support team to help you with this.
Save backup to a bucket supported by Scylla Manager.

For additional information:

Scylla Monitor ¶

Scylla Monitoring helps you monitor everything about your Scylla cluster. Scylla Support usually asks for your monitoring metrics when you open a ticket.

For information see:

Scylla Monitor

Configuration Management ¶

Caution

All configuration settings for all nodes in the same cluster should be identical or coherent.

Using tools such as Ansible, Chef, Puppet, Salt, or Juju are recommended.

See this article for more information.

Security ¶

Use the following guidelines to keep your data and database secure.

Enable Authentication
Create Roles for all users and use RBAC with or without LDAP (coming soon).
Use Encryption in Transit between nodes and client to node.
Refer to the Security Checklist for more information.

HA Testing ¶

HA testing in single DC - for example:

Shutdown one node from the cluster (Or scylla service if on the cloud) for 30 min.
Turn it back on.

HA testing in multi DC - for example:

Disconnect one DC from the other by stopping scylla service on all of these DC nodes.
Reconnect the DC.

Additional Topics ¶

Add a Node
Repair
Cleanup
Tech Talk: How to be successful with Scylla