Future-Proof Cloud Platforms 3 key factors for SaaS
The IT infrastructure of a company can be overwhelming. It consists of a greenhouse, large Cloud systems, an increasing number of micro-services and, in addition, the home office is bringing new interfaces. Mike Tria Atlassian has identified three key factors for future-proof Cloud platforms.
CompaniesSaaS vendors, it needs to rethink how they provide their customers with reliable and secure Cloud infrastructure available.
(© kras99 – stock.adobe.com)
SaaS vendors, it needs to rethink how they provide their customers with reliable and secure Cloud infrastructure available. Companies must create a new Standard in order to meet your customers ‘ expectations, because it is the only reliable systems allow for successful Service provision.
Conversely, however, this means that developers do more. As Head of Platform Engineering at Atlassian, the author was able to make these experiences with his own Team, and three key factors to filter out in order to future-proof, scalable, and reliable Cloud-build platforms.
1. Micro services
The heart of every SaaS Cloud platform, the complexity is. Complexity can be managed in two ways: centralized or distributed. A centralized System is a Monolith, in which the entire complexity of lives in a single System that provides a single interface to the outside world.
A distributed System is often built with the help of micro-services, in which the complexity in individual services will be split, which in turn communicate with each other. At first glance, the micro-services based System seems to be more fragile, since there are more interfaces, and thus more opportunities for error.
However, the reverse is the case: By the use of Service Level Objectives (SLOs) for each micro service and additional Alerts to relevant Teams a “defense-in-Depth is created” in the System. Some companies already use SLOs, but only in the outer layers of your system.
Atlassian, for example, relies on this defense-in-Depth and uses over 1,400 micro-services. These micro-services with attached SLOs, which are associated with Alerts, are the key to a reliable large-scale system.
2. Automation anywhere where it is possible
No longer would already make up five percent of the Issue Services your Job would be all the reliability there. By Automation in all sorts of Places can be ensured, however, that the reliability is almost 100 percent. A separate Tool monitors all existing micro-services, associated alerts in case of anomalies of individual SLOs and forwards them to the appropriate Teams so that they can take.
This Tool should continue to be an Incident Management System coupled to each Protocol is reported violation directly and can be immediately corrected. In addition, a micro-service System to effectively scale, which opens up eventually, other Engineering-resources for demanding Work.
3. An error budget to introduce
Thus, the reliability of Services can be properly assessed, should in the future be, an internal “error budget” introduced. This concept is used by modern Software companies for the past two to three years.
A big advantage of error budgets is that they give the team room. As long as the service is within its error budget, allows the Team to work as usual. Only if the error budget is exceeded, it should move the Team to an emergency tactics.
As an example: The access team is given the task that the Service must respond in 99.99 percent within a second. The number of minutes, where this number is not reached, then the error budget. Adopted from the calculation result, 52 minutes of the year are error budget. Something goes wrong and the Team needs two minutes to respond, that is within the budget, and requires no further action.
Previously treated Atlassian all errors the same, with the use of an error budget and SLOs is different – errors are reviewed and are now scalable. Atlassian’s all the Services to receive this classification by default.
Reliability is the Foundation for successful Service provision
Without a customer no Service and no reliable Service, no customer. However, in order to establish such services, come around, most companies is not a Revision of its internal Service structure. With a shift to micro-services, a Central platform for the management and exploitation of the error budget, you lay the ideal Foundation to the customers on an ongoing basis reliable, and scalable services.
* Mike Tria is Head of Platform Engineering at Atlassian.