To give you some background; the 2.4 release really focused on ensuring that we dealt with any scalability barriers detected in our stress tests. With v2.6 the focus has been on including even more extensive integration capabilities, to enable our customers to build innovative services and solutions.
We know that the “one size fits all” approach doesn’t work in the cloud marketplace, so an important goal for Abiquo is to enable our customers to offer a unique combination of services and solutions. With v2.6 we have enabled new integration capabilities, exciting enhancements to our outbound API, further back-up integration, workflow improvements and the opportunity to manage your public cloud through Amazon EC2.
Below is a review of these capabilities:
Imagine that you can subscribe all of your apps to an event bus, where you can detect Power On/Off or any other event defined by the Abiquo platform. Imagine that you can add information generated by Abiquo to maintain the information in your enterprise CMDB or your monitoring system, or your cloud customers can be subscribed to events (restricted by our roles and scopes security model that’s defined by you, of course) and can integrate to their own CMDB or service desk. This enables everyone to see the cloud information in the same way they can see the IT information in their local datacenter. Abiquo allows you to let your imagination run wild!
The Outbound API is the functionality that exposes all events generated by Abiquo and enables you to develop clients that can collect this information using long polling techniques, side events and so on.
To develop this capability, we have used our existing RabbitMQ system, publishing any event generated on Atmosphere, enabling anyone to build massive, scalable and real-time asynchronous applications using transports like WebSocket, Server Side Events and traditional Ajax Techniques.
Backup Integration – VM tags
As you may know, at Abiquo we want to free our customers from using any specific infrastructure stack and as a result we have added backup capabilities to our Outbound API.
As an orchestration layer, we also do not want to cause drastic upheaval to the underlying technology, with this in mind our Backup feature was created.
Today there are many dozens, or perhaps even hundreds, of backup solutions and with a single tool, there are different ways of operating depending on the DC capacity (for example, maybe you would not want to offer a user a backup at 2:00am if your DC infrastructure could not assure this SLA for all of your cloud users).
So why not create the interface and enable customers to integrate with their favorite solutions through the Outbound API? This is what our backup integration does:
- You can enable Backup at DC level and, using privileges, you can offer the capabilities to configure VM backup, should you wish to.
- You can offer hourly, daily, and weekly backups. Users can define the desired days or the defined hour
- Then this information is exposed in the Outbound API, enabling the development of plugins for the desired backup technologies
- Then these backup plugins can offer feedback to the user about the status of their requests through a simple API call, and the results are shown through the Abiquo GUI.
Many interesting things can be done with the Outbound API, and backup integration is just one of them.
Do you want to integrate the Deploy, Undeploy and Reconfigure processes with a third party system? Imagine that changes in the platform must be approved by the IT team. (Yes, that’s not cloud, but it can happen). Or imagine that you want to perform external validations that Abiquo currently does not cover because they are very specific to a certain market. Or maybe you want to apply some network configuration before deploying a virtual machine. For all of these cases and for an IT team with complex requirements, our workflow interface enables you to do that.
You will see some examples of this in future blogs in the following days but here’s a quick summary:
- The customer can enable workflow and define an endpoint where the Abiquo platform will publish the request to accept or reject a specific task
- When a user wants do deploy – as an example, the scheduler validates all Abiquo business rules and pre-selects a location to deploy the VM
- Instead of deploying directly, it publishes the request to the endpoint. The VM is given a “waiting for validation” status
- Then a third-party tool can receive this request. Here, customer imagination is key because you can connect with any system you like, such as a credit check system, finance system, service desk or just email
- When the external process is completed, you can accept or reject the task using a simple API call.
We have developed an improved hypervisor plugin model. Do you want to develop a RHEV plugin? Now you can. Do you want to develop an integration with vCenter to avoid connecting each ESXi separately? Now you can.
The new hypervisor plugin system, like the Storage one, enables you to quickly develop and easily plug in to the platform. In addition, the system incorporates and supports two modes: hypervisor (working directly with the hypervisor) or hypervisor manager (to abstract the connections with the hypervisor and interact with the virtualization system through the manager). Our Oracle VM plugin is a good example of the hypervisor manager mode, another example will be XenServer Pools. Do you want to create a connector to OpenStack? We want to hear from you!
Our Oracle VM and Hyper-V 2012 integrations are two examples where we have used this new plugin to extend our list of supported hypervisors.
But there’s more to v2.6!
With v2.6 you can now deploy to Amazon (manage your Amazon account and your on premise DC through a single pane of glass), v2.6 has improvements to the VM template library, v2.6 has JSON support in all API calls, and managing the limits of your NFS repository. Now it’s your turn.
A product download is available on our website. We look forward to hearing your feedback and to help you with your exciting integrations.