Budget about US$3.000 per direct IT staff for an AWS migration
AWS Migration and its challenges
Having more than one million users, and 90% of the total Cloud industry capacity in infrastructure and platform availability; the AWS or Amazon Web Services cloud platform is becoming a de facto standard. In order to properly capitalise on the benefits of running your workloads of data in the AWS cloud, many organisations already have, or at the very least are considering, an AWS Migration in order to properly manage their IT infrastructure. AWS is not however, all that simple to use. You will need training and new skills development.
With years of experience in Cloud migration, AWS migration and maintaining and running Cloud CRM software in order to optimise business processes, we can make a few observations about AWS.
AWS Migration and its Essential Benefits
Migrating to AWS will allow you to have access to a variety of complex cloud methodologies as well as different cloud hybrid models which are specifically designed to enhance your data handling process. Importantly AWS uses a best of breed tool and components approach, so you will be forced to properly understand, revise, re-code, document, organise, clean and extract your IS and corporate data. The embedded best of breed approach at all levels of the IT stack is a primary benefit of moving from your own corporate data centre to a market leading infrastructure.
Challenges Related to AWS Migration
Migrating to the AWS cloud is not an easy task to handle as there are a lot of tasks and details at every level of the stack, to be considered and taken into account. The components and levels within AWS can be complex and you will need to train your engineers, network admin staff and other IT resources in the various AWS layers. In other words your IT skills will be changed, improved, upgraded and exposed to leading edge ideas. You should budget US$3000 per individual, to retrain your staff.
Some challenges – a few amongst many:
- Throughput and speed of data transfer to the end user. AWS is not as fast as Google CE in data throughput. There are techniques to cache data, use Edge centers and other components within the AWS architecture to greatly improve the transfer speed of data. These methods are however, not that simple to implement [eg. Elasticache, MemCache]
- Transferring a large amount of information to the AWS cloud. There are a variety of tools which can manage the data migration from data import/export tools for smaller loads; to physical boxes which can handle massive terabyte loads [snowball for eg.]. You will need to be trained in these methods however in order to migrate the data properly.
- You will need the proper skills to set up your VM, manage your instance [database on AWS], create your AMI [machine images which copy the OS and apps], deploy the AMIs, customize them, secure your instances, create your IAM security models, backup your data, know the differences between S3, Glacier, RDS etc. and optimize your code to AWS…..
- You will need to understand how to audit, track, and monitor your system and usage in total and by end user.
- Architecture knowledge and the creation of hybrid models linking AWS to your corporate data stores, requires direct experience with the AWS stack, or at the very least a lot of training.
- Security for instance, can be layered in within AWS at each level of the stack, but you need to be aware of how to do this through the AWS console, what setups work, and the main issues around NACs, ports, firewall, user groups, IAM, permissions and data locking. For many who are new to AWS, this can be daunting. You can secure an EC2 instance for example, but end up blocking off some users from using this database process. Data left open will be prone to attack, and the complexity of the IAM model related to data storage and your EC2 instances can easily confuse engineers.
- There are many different sizes of instances that you can choose from and this range is quite complicated. It could mean that you end up paying a lot more for instance usage, than you need to, because you did not comprehend the differences between the various instance types.
- A common issue in AWS is not cleaning the web facing or data usage files properly. It is important that you can be sure of the files and resources that you will need in the future so that duplicate files are not stored and billed for. For example, version control can result in huge data sets being replicated with small net changes resulting in the entire file being copied [including the small net-new changes].
These are just some of the issues with AWS. To properly deploy a hybrid cloud model on AWS training will be necessary and you should budget about US$3000 per person to develop the requisite skills in deploying and managing your hybrid cloud application.
