SCALEWAY compute instances impacts

[demeringo]

Problem

As we did for AWS (https://boavizta.github.io/boaviztapi/Explanations/devices/cloud/) we should be able to retrieve the impacts of Scaleway instances.

Solution

The same implementation as AWS instances.

Alternatives

Additional context or elements

[Shillaker]

I should be able to look at this 👍

[da-ekchajzer]

To add a new provider :

• add the provider in data/archetypes/cloud/providers.csv : https://github.com/Boavizta/boaviztapi/blob/dev/boaviztapi/data/archetypes/cloud/providers.csv
• add a csv file (data/archetypes/cloud/scalway.csv) which contains all the instances' configuration in the right format :

id	manufacturer	CASE.case_type	year	vcpu	platforme_vcpu	CPU.units	CPU.core_units	CPU.name	CPU.manufacturer	CPU.model_range	CPU.family	CPU.tdp	CPU.manufacture_date	instance.ram_capacity	RAM.capacity	RAM.units	SSD.units	SSD.capacity	HDD.units	GPU.name	GPU.units	GPU.TDP	GPU.memory_capacity	POWER_SUPPLY.units	POWER_SUPPLY.unit_weight	USAGE.instance_per_server	USAGE.time_workload	USAGE.use_time_ratio	USAGE.hours_life_time	USAGE.other_consumption_ratio	USAGE.overcommited	Warnings	configuration.disk.units	configuration.disk.type	configuration.disk.capacity
a1.2xlarge	AWS	rack	2018	8	16	1	16	Graviton	Annapurna Labs	Graviton	Graviton	40	2018	16	16	2	0	0	0					2;2;2	2.99;1;5	2	50;0;100	1	35040	0.33;0.2;0.6	0		0

• If the CPU is not listed in https://github.com/Boavizta/boaviztapi/blob/dev/boaviztapi/data/crowdsourcing/cpu_specs.csv it is recommended to add it if you have the information. If not, the completion process will be used each time the instance is requested.

Feel free to ask any questions.

[da-ekchajzer]

Adding new cloud providers and new cloud instance should now follow this process : https://doc.api.boavizta.org/contributing/cloud_provider/

[Shillaker]

Yes it looks great, thanks @da-ekchajzer. I am adding a boavizta-dump command to this SCW + Boavizta PoC to generate the CSV files. Once that's done, I'll be able to copy-paste into a PR on this repo.

[Shillaker]

@da-ekchajzer, I see that in the server docs there is a field labelled CPU.threads with a description Number of vCPU per CPU. AFAICT this corresponds to the column labelled CPU.vcpu in server.csv, and so I'm pretty sure it's just a typo and should be CPU.vcpu in the doc.

However, can I just double check that this is the correct assumption? I assume that the column CPU.vcpu should be the max number of vCPUs allocated per CPU on that server type, and not the number of threads available on the CPU (which I guess is taken from cpu_specs.csv).

I also don't see any Ampere CPUs in the CPU specs. I'm happy to add them too, but can I check I'm not missing something?

[da-ekchajzer]

@da-ekchajzer, I see that in the server docs there is a field labelled CPU.threads with a description Number of vCPU per CPU. AFAICT this corresponds to the column labelled CPU.vcpu in server.csv, and so I'm pretty sure it's just a typo and should be CPU.vcpu in the doc.

Yes, you are right ! I just fixed it.

I also don't see any Ampere CPUs in the CPU specs. I'm happy to add them too, but can I check I'm not missing something?

You should indeed add the missing CPUs by following this doc: https://doc.api.boavizta.org/contributing/cpu/. I have just added the following note to the server contrib' page:

If CPU.name is set and the CPU is available in cpu_specs.csv, you do not need to fill in the other CPU attributes. The API will complete them based on the CPU.name.

However, can I just double check that this is the correct assumption? I assume that the column CPU.vcpu should be the max number of vCPUs allocated per CPU on that server type, and not the number of threads available on the CPU (which I guess is taken from cpu_specs.csv).

We assume that the max number of vCPU is equal to the number of thread (same value, different name depending on the context). If vCPU is not avaible in the server.csv, the API will complete vCPU with the number of threads taken from cpu_specs.csv based on the CPU name.

Is it a bad assumption ?

[Shillaker]

Awesome, thank you! Very clear.

Is it a bad assumption ?

No it's definitely sensible as a default. It's not always true though, so having the CPU.vcpu field in server.csv allows people to override it when it's not the case, which is perfect.