Optimized for spot instance utilization
Based on cloud region performance and internal AZ score
AI-based analytics
from OptScale
100% free,
no hidden costs
Hystax constantly analyzes spot instance availability, cloud region status & performance, price and metrics monitoring collected from Hystax OptScale customers to return the optimal configuration for your CI/CD job or other R&D resources.
You provide an expected job duration, desired cloud, region and flavors for new VMs as an input parameters, the API returns what region, availability zone, instance flavours and types to use for your task or CI/CD job.
US and European regions of AWS are supported by the API.
Spot instances are 1.5-4x more cost efficient than on-demand instances. Depending on cloud type, desired region and instance flavors we try to find the best configuration from price perspective and spot instance utilization.
We constantly monitor spot instance revocation by clouds and return flavors and instance types with the lowest possibility of termination in a specific region and availability zone.
The internal AI-based engine grabs price data from all the cloud regions and returns the most cost-efficient options.
API uses cloud health monitor and availability zone score from Hystax OptScale to exclude regions experiencing troubles with performance and latency. Based on your desired cloud and regions, the API returns only the healthy regions and availability zones close to the desired region to avoid data locality and latency issues.
The API is free with no hidden costs or plans. You can use one token for your team or entire organization for an unlimited number of R&D workloads, CI/CD jobs or other provisioning tasks. To prevent DDoS attacks we limit 5 API calls per second for one token.
You can always upgrade to Hystax OptScale to get cloud arbitrage, cost management and migration capabilities based on your cloud usage patterns and cloud health monitor.
The API takes JSON as an input parameter. There are the following JSON parameters you need to provide with each call:
{
"allow_nearby_regions": true,
"alternative_cloud": "aws",
"cloud": "aws",
"machines": [
{
"ram": 4,
"vcpu": 2
},
{
"projected_duration": 4,
"ram": 4,
"vcpu": 2
}
],
"optimize_by": "price",
"projected_duration": 24,
"region": "us-east-1"
}The API returns JSON as an output with the following parameters:
[
{
"region": "us-east-2",
"cloud": "aws",
"availability_zone": "us-east-2c",
"instancetype": "t4g.medium",
"type": "on-demand",
"notes": "cheapest region, best suited and cheapest flavor"
},
{
"region": "us-east-2",
"cloud": "aws",
"availability_zone": "us-east-2b",
"instancetype": "a1.large",
"type": "spot",
"notes": "spot due to duration, cheapest region, best suited and cheapest flavor"
}
]* Spot instances may be terminated by a public cloud at any time with a prior 5 minute notice. The API returns a configuration with the highest possibility of completion in a timeframe (job duration) provided as an input parameter, but cannot guarantee 100% success ratio. Handle the termination by either restarting them or completing with an error.
For more details, please refer to the article: ‘Spot instances for CI/CD jobs. Yes or No?’
A full description of OptScale as a cloud arbitrage platform – product features, benefits and functionality
Great critical insights on hybrid cloud management benchmarks, trends and best practices
