Plan for production

Elasticsearch Add-On for Heroku supports a wide range of configurations. With such flexibility comes great freedom, but also the first rule of deployment planning: Your deployment needs to be matched to the workloads that you plan to run on your Elasticsearch clusters and Kibana instances. Specifically, this means two things:

With Elasticsearch Add-On for Heroku, your deployment can be spread across as many as three separate availability zones, each hosted in its own, separate data center. Why this matters:

We recommend that you use at least two availability zones for production and three for mission-critical systems. Just one zone might be sufficient, if your Elasticsearch cluster is mainly used for testing or development and downtime is acceptable, but should never be used for production.

With multiple Elasticsearch nodes in multiple availability zones you have the recommended hardware, the next thing to consider is having the recommended index replication. Each index, with the exception of searchable snapshot indexes, should have one or more replicas. Use the index settings API to find any indices with no replica:

GET _all/_settings/index.number_of_replicas

Moreover, a high availability (HA) cluster requires at least three master-eligible nodes. For clusters that have fewer than six Elasticsearch nodes, any data node in the hot tier will also be a master-eligible node. So, this can be achieved by having hot nodes (serving as both data and master-eligible nodes) in three availability zones, or by having data nodes in two zones and a tiebreaker (will be automatically added if you choose two zones). For clusters that have six Elasticsearch nodes and beyond, dedicated master-eligible nodes are introduced. When your cluster grows, it becomes important to consider separating dedicated master-eligible nodes from dedicated data nodes.

The data in your Elasticsearch clusters is also backed up every 30 minutes, 4 hours, or 24 hours, depending on which snapshot interval you choose. These regular intervals provide an extra level of redundancy. We do support snapshot and restore, regardless of whether you use one, two, or three availability zones. However, with only a single availability zone and in the event of an outage, it might take a while for your cluster come back online. Using a single availability zone also leaves your cluster exposed to the risk of data loss, if the backups you need are not useable (failed or partial snapshots missing the indices to restore) or no longer available by the time that you realize that you might need the data (snapshots have a retention policy).

Knowing how to scale your deployment is critical, especially when unexpected workloads hits. Don’t forget to check your performance metrics to make sure your deployments are healthy and can cope with your workloads.

Scaling with Elasticsearch Add-On for Heroku is easy:

Refer to Sizing Elasticsearch: Scaling up and out to identify which questions to ask yourself when determining which cluster size is the best fit for your Elasticsearch use case.