Instance types

Instance types on Wavestack

Every compute instance running in Wavestack has a specific type, or flavour in OpenStack parlance. The type defines various combinations of instance attributes such as number of virtual CPU cores, amount of memory, network or storage capacity.

You can choose between a wide variety of instance types on Wavestack, which allows you to tailor your selection to your specific needs.

Execute the following command to list all currently available instance types/flavours:

❯ openstack flavor list

We follow the Sovereign Cloud Stack naming standard. If you want to learn more about it and understand the logic behind these names, you can read more about it in:

SCS - Flavor Naming Standard v3.

General Purpose

General purpose instances provide a balanced ratio of virtual CPU cores and memory. They are a good choice for a wide variety of use cases unless your workloads have more specific requirements.

Typical workloads:

Web servers
Business intelligence applications
Containerised services
Development and test environments

	vCPU	Memory	Disk	Disk Type	Average Bandwidth	Peak Bandwidth
SCS-2V-8	2	8	0	rbd_fast	0.512	1
SCS-2V-8-20	2	8	20	rbd_fast	0.512	1
SCS-4V-16	4	16	0	rbd_fast	1.25	2.5
SCS-4V-16-50	4	16	50	rbd_fast	1.25	2.5
SCS-8V-32	8	32	0	rbd_fast	1.25	2.5
SCS-8V-32-100	8	32	100	rbd_fast	1.25	2.5
SCS-16V-64	16	64	0	rbd_fast	2.5	5
SCS-16V-64-200	16	64	200	rbd_fast	2.5	5
SCS-32V-128	32	128	0	rbd_fast	2.5	5
SCS-32V-128-500	32	128	500	rbd_fast	2.5	5
SCS-64V-256	64	256	0	rbd_fast	5	10
SCS-64V-256-1000	64	256	1000	rbd_fast	5	10

Compute Optimised

Compute optimised instances are particularly well suited for compute bound workloads. We recommend to pick these if you your workloads are processing heavy, but do not require that much memory at the same time.

Typical workloads:

High-performance web servers
Scientific computation
Batch processing
Image or video processing
High-performance computing (HPC)

	vCPU	Memory	Disk	Disk Type	Average Bandwidth	Peak Bandwidth
SCS-8V-16-50	8	16	50	rbd_fast	1.25	2.5
SCS-16V-32	8	16	0	rbd_fast	2.5	5
SCS-16V-32-100	16	32	100	rbd_fast	2.5	5
SCS-32V-64	32	64	0	rbd_fast	2.5	5
SCS-32V-64-200	32	64	200	rbd_fast	2.5	5
SCS-64V-128	64	128	0	rbd_fast	5	10
SCS-64V-128-500	64	128	500	rbd_fast	5	10

Memory Optimised

Memory optimised instances are particularly suitable for memory bound workloads that benefit from fast access to large data sets.

Typical workloads:

Database servers
In-memory data stores (e.g. redis)

	vCPU	Memory	Disk	Disk Type	Average Bandwidth	Peak Bandwidth
SCS-1V-8	1	8	0	rbd_fast	0.512	1
SCS-1V-8-20	1	8	20	rbd_fast	0.512	1
SCS-2V-16	2	16	0	rbd_fast	0.512	1
SCS-2V-16-50	2	16	50	rbd_fast	0.512	1
SCS-4V-32	4	32	0	rbd_fast	1.25	2.5
SCS-4V-32-100	4	32	100	rbd_fast	1.25	2.5
SCS-8V-64	8	64	0	rbd_fast	2.5	5
SCS-8V-64-200	8	64	200	rbd_fast	2.5	5
SCS-16V-128	16	128	0	rbd_fast	2.5	5
SCS-16V-128-500	16	128	500	rbd_fast	2.5	5
SCS-32V-256	32	256	0	rbd_fast	5	10
SCS-32V-256-1000	32	256	1000	rbd_fast	5	10
SCS-1V-16	1	16	0	rbd_fast	1.25	2.5
SCS-1V-16-50	1	16	50	rbd_fast	1.25	2.5
SCS-2V-32	2	32	0	rbd_fast	1.25	2.5
SCS-2V-32-100	2	32	100	rbd_fast	1.25	2.5
SCS-4V-64	4	64	0	rbd_fast	2.5	5
SCS-4V-64-200	4	64	200	rbd_fast	2.5	5

Disk I/O Optimised

Disk I/O optimised instance are equipped with fast local storage and are well suited for workloads that require frequent disk access.

Typical workloads:

Key-value stores (e.g. etcd)
Kubernetes controlplane nodes

Please note that local storage is ephemeral, which means that data stored on these volumes is lost if the instance is destroyed.

Data will also not be migrated or evacuated when maintenance tasks are performed.

Please ensure that your workloads can tolerate this behaviour when utilising instances with local storage or opt for a different instance type.

	vCPU	Memory	Disk	Disk Type	Average Bandwidth	Peak Bandwidth
SCS-2V-8-20s	2	8	20	ssd	1.25	2.5
SCS-4V-16-50s	4	16	50	ssd	1.25	2.5
SCS-8V-32-100s	8	32	100	ssd	2.5	5
SCS-16V-64-200s	16	64	200	ssd	2.5	5
SCS-32V-128-500s	32	128	500	ssd	5	10
SCS-64V-256-1000s	64	256	1000	ssd	5	10
SCS-1V-8-20s	1	8	20	ssd	512	1
SCS-2V-16-50s	2	16	50	ssd	1.25	2.5
SCS-8V-64-200s	8	64	200	ssd	2.5	5
SCS-4V-32-100s	4	32	100	ssd	2.5	5
SCS-16V-128-500s	16	128	500	ssd	2.5	5
SCS-32V-256-1000s	32	256	1000	ssd	5	10

AI/ML Optimised

AI/ML optimised instances give you access to GPU resources and are well suited for high-performance computing workloads.

Instance types with NVIDIA A30 Tensor Core GPUs are currently only offered in the muc-a availability zone.

	vCPU	Memory	Disk	Disk Type	Average Bandwidth	Peak Bandwidth	VRam
SCS-16V-64-500s_GNa-14h	16	64	500	ssd	2.5	10	6

Please get in touch with our support if you require differently sized GPU based instances!

Budget Optimised

These instances are offered at a lower cost and are a good choice for non-critical workloads.

Typical workloads:

Non-critical processes
Long-running batch processes
Development or test environments

	vCPU	Memory	Disk	Disk Type	Average Bandwidth	Peak Bandwidth
SCS-1L-1	1	1	0	rbd_fast	0.512	1
SCS-1L0	1	1	5	rbd_fast	0.512	1
SCS-1L1	2	1	0	rbd_fast	0.512	1
SCS-1L2	2	1	50	rbd_fast	0.512	1
SCS-1L3	4	2	0	rbd_fast	0.512	1
SCS-1L4	4	2	10	rbd_fast	0.512	1
SCS-1L5	8	4	0	rbd_fast	1.25	2.5
SCS-1L6	8	4	20	rbd_fast	1.25	2.5
SCS-1L7	16	8	0	rbd_fast	1.25	2.5

Notes

Storage Types / IOPS

Performance attributes differ between storage/disk types, which are summarised in the table below:

	Average IOPS	Burst IOPS	Average Throughput	Burst Throughput
rdb_fast	1500	10000	0.25	0.5
ssd	10000	10000	0.75	0.75

Units

Information in this document is provided in the following units:


Memory	GiB
Disk	GiB
Bandwidth	Gbit/s
Disk Throughput	GiB/s