I've been unexpectedly brought into a meeting set up by our Director with a third-party company that specializes in Cloud migrations. They're going to discuss data center consolidation and reducing total cost of ownership (TCO) for our systems. From what I understand, this company tried this approach before I joined and it wasn't cost-effective then, and I doubt it will be now. However, I'm not an expert in Azure, so I need some advice.
Our on-premises equipment is less than three years old and has plenty of resources available; most of it is VMware, which can be quite expensive when we renew licenses. Also, we use Office 365 under a national shared tenancy, so I don't expect any savings there either.
We run a wide variety of applications (around 200+) which typically consist of two web servers, two application servers, and one SQL server, either standalone or in clusters for specific applications. I'm concerned that if we do a partial migration, like just SQL or application servers, it could increase latency since it would involve Internet links. Would we need to migrate all components of each application at once?
Any insights or advice as I prepare for this discussion?
5 Answers
If they're promising savings, ask them straight up for a written guarantee. You’ll likely see them squirm! It's a classic sales tactic, and there's so much scope for unexpected costs in cloud setups that you shouldn’t trust their word alone.
Just throwing this out there, but cloud migrations often only see savings if you go with a full PaaS or SaaS model, rather than IaaS. If any of this is handled poorly, you could see costs skyrocket instead of diminish. A lot of businesses have gone back to on-prem after diving into the cloud, especially when dealing with hybrid setups. It’s important to have a deliberate strategy in place; otherwise, the benefits will be overshadowed by additional costs and performance issues. Take a look at some case studies on cloud repatriation to see what others have experienced.
You're right to be skeptical about the cost savings. If your management has been treated to nice lunches, they're likely to be influenced into thinking the cloud is a win. To combat that, draft a comprehensive cost analysis that compares your current on-prem expenses and potential future costs of running everything on the cloud over the next 5 to 10 years. Include all costs like VMware licenses and extended support for your current servers. It's also worth noting that your applications may experience increased latency if you don't migrate everything in sync. But, if they're adaptable, you might be able to mix and match components between on-prem and cloud. Just ensure the coding is efficient to mitigate any delays.
Have you considered alternatives like migrating from VMware to Proxmox instead of Azure? Sometimes looking at other virtualization options can yield better cost efficiency. There are definitely instances where cloud solutions just aren't the best choice financially.
You really need to be careful with a lift-and-shift approach in Azure. It's often more expensive than keeping things on-prem. One key thing to remember is that Azure bills based on usage. If you’re just moving your current setup without rethinking your operations, you might end up incurring unexpected costs. The way to maximize savings is to shift to services like SQL as a service, which eliminates the need for managing clusters. You might want to run a cost estimate using Azure's calculator based on your existing infrastructure to make your case more tangible. Oh, and I did a similar migration and found that if you separate SQL from app servers, performance really tanked. It's better to migrate both together or leave them on-prem.
Related Questions
Can't Load PhpMyadmin On After Server Update
Redirect www to non-www in Apache Conf
How To Check If Your SSL Cert Is SHA 1
Windows TrackPad Gestures