Elasticsearch Add-On for Heroku Azure default provider instance configurationsedit

Elasticsearch data nodes are available in these configurations:

  • highio - Optimized for balanced RAM/vCPU/disk ratios and performance.
  • highstorage - Optimized for cost effective storage.
  • datafrozen - Optimized for frozen data storage, acts as a cache for the frozen tier.
  • highcpu - Optimized for higher CPU performance compared to highio types.
  • highmem - Optimized for lower cost with lower storage ratio.

The following table contains the specific hardware for Azure. For GCP and AWS, we are introducing a new set of instance configurations.

Instance configuration Provider Preferred Instance Type or Machine Configuration1 Storage Type1 Disk:Memory Ratio2 vCPU/RAM Ratio

highio

Azure

l32sv2

NVMe

30:1

0.131

highstorage

Azure

e16sv3

Standard Managed Disks (HDD)

160:1

0.133

datafrozen

Azure

lsv2

NVMe

31600:1

0.133

highcpu

Azure

d64sv3

Standard SSD

8:1

0.258

highmem

Azure

e32sv3

Standard SSD

2:1

0.131

Additional instancesedit

The following table contains the specific hardware for Azure. For GCP and AWS, we are introducing a new set of instance configurations.

Instance configuration Provider Preferred Instance Type or Machine Configuration1 Storage Type1 Disk:Memory Ratio2 vCPU/RAM Ratio

apm

Azure

e32sv3

Standard SSD

2:1

0.131

appsearch or enterprisesearch

Azure

d64sv3

Standard SSD

2:1

0.258

ccs

Azure

e32sv3

Standard SSD

2:1

0.131

coordinating

Azure

d64sv3

Standard SSD

8:1

0.258

kibana

Azure

e32sv3

Standard SSD

2:1

0.131

master

Azure

e32sv3

Standard SSD

2:1

0.131

ml

Azure

d64sv3

Standard SSD

2:1

0.258

1 Preferred instance and storage types are subject to provider availability. To learn more about our provider instances, see AWS, Azure, and GCP reference pages.

2 Ratios are estimations.

3 This is the amount of searchable storage in the frozen tier snapshot. The cached storage volume is a lower value than the searchable storage volume, ranging from about 30:1 to 90:1 depending on the hardware type.

JVM Heapedit

Memory sizes ensure efficient hardware utilization and might not scale to the power of two (n2). For sizes above 58, 60 or 64 GB (depending on instance type), we create multiple instances or nodes to ensure efficient JVM heap sizes. For example: If you provision a deployment with a 128 GB Elasticsearch cluster, two 64 GB nodes get created. To learn more about why we offer these JVM heap sizes, see Heap: Sizing and Swapping.