When it comes to suggesting the most cost-effective way to start/stop Virtual Machines in order to reduce cloud costs, some businesses baulk at the idea of using scheduling software. “Why pay for what we can do manually?” and “why buy when we can build?” are two of the most frequently-used reasons for not considering scheduling software. However, both the manual process and the script-writing process have potential issues associated with them that are difficult to overlook.
The Potential Issues of Starting/Stopping Instances Manually
The potential issues of starting/stopping instances manually are reliability, inaccuracy and cost (yes – cost). The reliability issue is easiest to explain. Although instructing developers to switch off non-production instances when they are not using them will likely be effective for a couple of weeks, it is unlikely to be a long-term solution to reducing cloud costs. Developers forget, bad habits creep in, or they fail to switch them off the right way so you are still paying for provisioned resources.
The “accurate” way of starting/stopping instances manually is to start/stop each instance through the AWS, Azure or GCE portal rather than via the remote desktop. The process can be time-consuming (which costs money) and, if you need your instances to retain the same IP address, you may also have to pay for a static IP address; as stopping an instance through the AWS, Azure or GCE portal results in all the resources allocated to it – including the IP address – being reallocated elsewhere.
The Potential Issues of Writing Start/Stop Scheduling Scripts
The potential issues of writing start/stop scheduling scripts are cost, inflexibility and reliability (again). This time cost is the easiest issue to explain. Reassigning developers from their core projects to write scheduling scripts costs money. Unless you pay your developers a very small salary, the cost of writing start/stop scheduling scripts will be more than the cost of implementing scheduling software, plus by reassigning developers from their core projects, you are risking the late delivery of their projects.
Inflexibility only becomes an issue if you have developers who work irregular hours. In these circumstances, each scheduling script would have to be edited as access to each instance is required in order to switch it on. Not only is this process time-consuming (which costs money), but there is the chance developers will forget to re-edit the scheduling script in order to switch the instance off when they have finished using it – hence the potential issue of reliability.
Overcoming Cost, Inaccuracy, Inflexibility and Reliability
Provided you do not implement an all-singing and all-dancing cloud management solution that you will never use of the features of, scheduling software typically pays for itself within two months. As it automates the process of starting/stopping instances, there are no issues about doing it the right way or wrong way (or for the wrong times); and, whereas manual and scripting processes lack controls, scheduling software has all the tools required to increase governance and accountability.
With regard to overcoming inflexibility, once the schedules for starting/stopping instances have been applied, they can be interrupted with the click of a mouse so developers can quickly and easily access them. Scheduling software also comes with safeguards in place so that, if developers forget to reactivate the instances´ schedules, the instances automatically switch off after a period of inactivity – the automatic process overcoming the issue of reliability.
Starting/Stopping Instances Using ParkMyCloud
ParkMyCloud is a lightweight SaaS app that focuses on automating the process for starting/stopping instances on AWS, Azure and GCE. In order to automate the process, system administrators create a permissions-limited credential for our app and it allow to conduct a discovery of the business´s cloud accounts. When the discovery is finished, ParkMyCloud will display all the business´s instances across a single-view dashboard – indicating those that are suitable for scheduling.
System administrators simply selected the highlighted instances and assign a parking schedule to them. A selection of preconfigured schedules is available to choose from, administrators can create their own schedules via ParkMyCloud´s intuitive GUI, or – if your development teams work irregular hours – permanently park the highlighted instances and set up role-based access controls so developers can interrupt the schedules when they need access to their non-production instances.
Try Our Scheduling Software for Free
If you acknowledge that the manual and scripting processes for starting/stopping instances have potential issues, and you would like to compare your current process against the benefits of scheduling software, we invite you to take a free trial of ParkMyCloud. The trial will give you the opportunity to experience ParkMyCloud´s ease of use in your own environment and see how much you could save by automating the process for starting/stopping instances.
There is no obligation on you to continue using our service once your trial has concluded, but we should point out that ParkMyCloud has a free pricing tier – and we do not know how it is possible to get more cost-effective than that. To start your free trial, simply click on the “Try it Free” button. Alternatively, if you have any questions about starting/stopping instances with ParkMyCloud, do not hesitate to contact us. Our friendly Customer Support team will be happy to answer your questions.