Cloud Infrastructure Optimized for Production Services
Successful PaaS, SaaS & ISVs choose CloudSigma
Whether you are transitioning an existing product to the cloud or building out an entirely new service, which cloud you choose can be a critical success factor. How fast and easily can you get up and running? What about scaling up if the service proves successful? What level of service performance & availability is achievable? You need a cloud that can deliver on these fronts and more. It’s why an increasing number of services used today run on CloudSigma.
Migration & Time to Market
Migration and set-up are one-off sunk cost investments that are cloud specific; minimizing them as much as possible therefore makes sense. Every change you make to your set-up to meet a proprietary requirement of a cloud is vendor lock-in. Avoiding proprietary customizations is definitely optimal. The less time you spend getting things up and running in the cloud, the quicker your service will be production ready and the more agile you’ll be in the future to adapt quickly to changing customer needs.
What to Ask
- What changes do I need to make to standard operating systems to work in your cloud?
Things to Look For
- Compatibility with standard operating systems and drive formats. The more changes you’ll have to make, the more investment in dead time migrating you’ll face.
Our Cloud Solution
- What interfaces are available for importing and exporting my data?
- A simple and preferably free ability to upload and download drive images from your cloud account in a usable format and within a reasonable time frame.
- Upload & download full drive images directly to/from our webapp or over our API. It’s free, immediate and simple. We accept simple RAW format drive images which you can convert to or from very easily.
- How does your server infrastructure differ from standard topologies?
- Adoption of standard models for servers, drives & networking. The more a cloud uses ‘new’ types of implementations, the more proprietary the set-up and the more work you’ll have setting up your cloud infrastructure.
- You build your servers just like physical ones; size your CPU & RAM, create and mount drives to the server & then configure your public and private networking. Zero new training required; it’s completely straightforward.
“First I should say that your custom image loading is AMAZING and extremely well implemented through the uploading of images. VERY nicely done. This is very important for my company, and your method is far easier than alternatives.”Leading North American ISV
Scaling Up & Automation
You are live but how do you ensure your service remains responsive? That’s where the ability to scale up and down becomes very important. Likewise, the more you can automate your service the faster changes will happen and the less time you’ll spend maintaining your cloud infrastructure. Those time savings will allow your developers and IT department to concentrate on higher value tasks like improving your core product not worrying about if you have enough web servers!
What to Ask
- What are your provisioning times for infrastructure components such as servers, drives and networks?
Things to Look For
- The faster the provisioning times the better. When you’re in production if you need resources you’ll need them as fast as possible. Even for testing, keeping developers waiting isn’t efficient.
Our Cloud Solution
- We offer instant infrastructure provisioning. New server creation takes less than thirty seconds for a fully functioning cloud server you can login to. Need a new private network? Under five seconds. You get the idea.
- What API coverage of features do you have?
- In the cloud, automation is key especially in creating a responsive service that always performs. Having true feature coverage on the cloud API is critical.
- What services integrate onto your API?
- Integration into external tools is going to make life a lot easier for your developers. Whether it’s a private cloud stack like OpenStack and Abiquo or a cross cloud driver like Apache Libcloud, the wider the tool-set available, the more orchestration and management options you’ll have to choose from.
- As a leading public cloud provider we have integrations available for OpenStack, Ubuntu, CoreOS, jclouds, libcloud, fog.io and many more. We are steadily adding to these over time to give our customers the widest choice of how they manage their cloud infrastructure. Check out our wide range of external integrations in various languages to speed your developers’ day to day work.
“Creating and cloning of VMs is a very fast process. About 2 minutes to start 25 already created machines. There are many options to optimize and change the specifications of a VM – Processors, RAM, Storage, Network interface, which is very handy as rendering products usually need a lot of resources – the more the better.”Top Three Global Rendering Company
Achieving Cost Efficiency
Whilst cost savings aren’t the greatest benefit to using cloud infrastructure, deploying efficiently is clearly important. The difference between a profitable and competitive service and a lost making one can come down to purchasing efficiency of underlying infrastructure. Cost efficiency comes from a combination of granular resource sizing that tightly fits your requirements and a cloud billing system that closely matches and charges for only your actual usage over time.
What to Ask
- What billing period do you use to measure my resource usage?
Things to Look For
- The closer a cloud bills you to your actual usage level the better your cost efficiency. A shorter billing period is therefore better in order to catch infrastructure changes sooner.
Our Cloud Solution
- We track customer resource consumption on a short five minute billing cycle. We expose the full transaction log to customers via the webapp and API so they can easily audit and track their billing and usage patterns.
- Can I freely size, resize and combine resources?
- Look for greater flexibility. The more control and flexibility you have over resource sizing and configuration, the closer you can provision your cloud resources to your actual resource requirements.
- We offer ‘perfect provisioning’. The ability to freely size servers and storage as you require. You can combine resources within a wide minimum and maximum range without restriction.
- Do you charge for servers that are shut down?
- Look for a cloud that doesn’t charge for resources when not in use. When running a service you are going to have a load curve; in other words service usage will vary over time. The cloud gives you elasticity so you can bring up and shut down infrastructure in a responsive way. Using a cloud that reflects that varying usage is critical to achieve cloud purchasing efficiency.
- We charge customers based on aggregate consumption of each resource i.e. CPU, RAM, storage etc. You can purchase subscriptions for your 24/7 requirement level and freely ‘burst’ whenever you need more. In conjunction with our five minute billing cycle it means you can create an ideal combination of long term purchasing with short term ad hoc usage.
“We settled on working with CloudSigma in their Zürich data centre. CloudSigma provided the reliability, capacity and flexibility required by Deutsche Börse at a competitive rate.”Stephen Watling, Deutsche Börse
Reliability, Transparency & Support
A service can only be as reliable as the infrastructure it depends on. Choosing a cloud hosting partner that delivers high uptime and reliability plus a professional and transparent approach to incident handling will ensure you build a successful service run out of the cloud.
What to Ask
- What high availability and redundancy features does your cloud offering have?
Things to Look For
- Service availability is extremely important, customers are very intolerant of downtime. A cloud that let’s you achieve true high availability with components separated and in fail-over will support you towards that goal.
Our Cloud Solution
- Our core networking and switching is all highly redundant. Additionally we over ‘avoid’ functionality on both storage and compute to enable customers to explicitly separate key components that might be exposed to single points of failure.
- What availability level do you offer via your cloud service level agreement?
- A 99.9% availability level in a service level agreement might sound high but it still represents almost nine full hours of downtime per year without any compensation. The higher the availability level, the better.
- Our service level agreement offers a 100% availability level plus a sub-1ms network latency guarantee within our cloud. It means any outage over 15min will accrue SLA credits and more importantly is painful to us as well as our customers. It gives us a strong incentive to invest in redundancy and avoid outages which our record shows.
- Where can I view your service track record and incident history?
- Every provider will have incidents from time to time, some will be transparent about it, others less so. Information empowers you to manage your services’ customers better. The more information you have and the more timely that information the better.
- We post all customer affecting service incidents publicly at status.cloudsigma.com. Anyone can subscribe to receive alerts of incidents and maintenance events.
- How are incidents managed and communicated to your customers?
- If your service is down or degraded from the cloud, how your cloud service provider handles an incident can make the difference between you keeping or losing your end users. Look for a professional, transparent approach to incident handling. Direct access to support personnel is also desirable.
- We offer full incident management. We proactively contact all customers affected by an incident then provide regular updates thereafter. Customers receive an alert once an incident is resolved and we provide a full root cause analysis once available. We offer 24/7 email ticket and live chat support free to all our customers.
“Using CloudSigma has enabled us to increase our level of service for our own customers since we are now able to spin up resources much more quickly in the cloud with greater scalability, reliability and performance.”Patrick Rodies, CTO, Preview Networks