5 ms·

The problem is that there is always going to be a new Windows Server release, plus a new SQL Server release. Even if you keep your licenses, are you going to st

The problem is that there is always going to be a new Windows Server release, plus a new SQL Server release. Even if you keep your licenses, are you going to stay on that version for the lifespan of the system? Likely not. Then you start to pay. For the database, the OS plus SQL Server Standard is still going to be around $15K (this was around 4-5 years ago when I was involved with this). Now this is just one instance. Multiple that by multiple servers and instances and you will reach upwards of 6 figures for your costs. I guess the numbers work out differently for cloud but I don't imagine they would be that far off. For example, if my Azure costs are $10K a month, from that cost is likely $2K-$3K the Microsoft tax. Keep scaling your VMs and servers and that margin starts to look a bit less palatable. If you know the limits of your business, then it's probably not an issue. I think the sweet spot would be to use Postgresql for the database and Microsoft for the rest, as most of the cost is SQL Server for systems that aren't in the full Microsoft eco-system (BizTalk, SharePoint,etc.)

ktavera 12y ago
Licensing costs are not really an issue when you're on Azure. If your app ran fine under the BizSpark program subsidy of $150/month of azure credit then when you're out of the program you'd just have to pick up the $150/month hosting cost, not suddenly have to pay $15k+ for a MSSQL or Windows Server license.
- avenger123 12y ago
  I know that. I didn't make it explicit in my comment. The Microsoft tax I mentioned with Azure is the licensing costs. Microsoft isn't just going to give out SQL Server for free and just charge for the VMs and bandwidth. There is premium pricing for SQL Server and BizTalk and custom licensing pricing for SharePoint and Oracle software. Being on Azure doesn't mean you get a break on the licensing costs. They just amortize the cost over the hour the VM is up. EDIT: Fair point on the hosting cost being known and acceptable right from the start. I also don't mean to imply that you are not aware of the licensing costs, just wanted to make it clear. For real businesses, at some point it makes more sense to build their own "private cloud" and at this point these issues become relevant but for most start-ups, it's likely just not going to be an issue.
  - pistle 12y ago
    While SQL Server has and will continue to be an excellent technical option, the costs can be a drag. That said, the work various .net-related people have out or are working on have the very real availability or promise to lift that dependency, if not provide portability, with DB choice. See RavenDB (document DB) & biggy (Postgres 1st class). Supporting this for more traditional takes on relational store is that there is increasing experience in the .net world for applying coding patterns that can help projects pivot/abstract out the data tool dependencies. If you have a sensitivity in this area, there are options which, when calculated with the ancillary benefits of applying existing .net experience and the great MS free + cloud tools, would pan out very well for many problem spaces. Obviously, YMMV, but if you are doing a wide spectrum of cloud + mobile, Azure is very compelling and Xamarin can become nearly as large a burden cost-wise. That total burden still can be extremely competitive and the ramp for existing .net shops a pretty clear winner. But being a C# developer has never been so promising - and it wasn't ever that awful to begin with. If anything, the problem of coming from the old MS-dev platform-planning and having to understand the range of options and optimizing for costs around those is the weak spot.
viggity 12y ago
a 150 GB business Azure SQL instance is $225/month, sounds like a steal to me.
ryanjshaw 12y ago
> are you going to stay on that version for the lifespan of the system? Likely not. Not sure why you say that -- if it works, it works. I've just spent the last year working on a multi-million dollar project budget building a new system on a SQL Server 2005 cluster. Why? Because there was no need to budget for new hardware. (SQL Server 2008 R2 would have been a big step up, after that not really worthwhile upgrades, in our case.) > For example, if my Azure costs are $10K a month, from that cost is likely $2K-$3K the Microsoft tax. Keep scaling your VMs and servers and that margin starts to look a bit less palatable. Let's say my hosting costs are 10k p.m, 120k p.a. and represent 15% of revenue (this is just how my business works - if I need that much SQL power I'm doing a hell of a lot of business). 120/0.15 = 800k in revenue. Keeping things really simple, let's say my gross margin is 50% and corp tax is 20%. That leaves the business with a nice 320k profit. Sure I don't like spending 4.5% of my revenue on MS licensing, but at this point I can quit my job and port over to Postgresql (or more likely pay somebody else to). The point is that nobody wants to give away 4.5% but if it's the line between success and failure then you probably have bigger issues with your business model/environment. Don't get me wrong - for my current side project I'd love to use Postgresql as the backend and .NET for the rest but the problem with Postgresql is two-fold and really specific to me: 1. I'd spend a lot of my time learning a new set of tools for zero benefit when time is my most expensive resource right now. I'm very familiar with T-SQL and the MSSQL platform (even simple things like having MERGE available has saved me lots of time given my problem domain, nevermind the ability to reuse a lot of code I've written and optimised over the past 10 years and knowledge of various issues and errors I will encounter). Porting in 3 years will be easier because we'll have a working system to port from, and the risk to the project will be less because we aren't doing two things at once (learn a new platform + solve a new problem -- only the later is going to make me money). 2. I plan on using some of the MS BI tools for the first time as part of this project -- this represents quite valuable experience to have as a .NET developer/solution designer; if I go the Postgresql route I will spend a lot of time exploring and evaluating open source options with zero benefit to my day job career progression over the next 3 years - I don't have the luxury of neglecting my day job while I work on side projects, so I need to be careful I don't lose value
- avenger123 12y ago

ktavera 12y ago

Licensing costs are not really an issue when you're on Azure. If your app ran fine under the BizSpark program subsidy of $150/month of azure credit then when you're out of the program you'd just have to pick up the $150/month hosting cost, not suddenly have to pay $15k+ for a MSSQL or Windows Server license.

avenger123 12y ago
I know that. I didn't make it explicit in my comment. The Microsoft tax I mentioned with Azure is the licensing costs. Microsoft isn't just going to give out SQL Server for free and just charge for the VMs and bandwidth. There is premium pricing for SQL Server and BizTalk and custom licensing pricing for SharePoint and Oracle software. Being on Azure doesn't mean you get a break on the licensing costs. They just amortize the cost over the hour the VM is up. EDIT: Fair point on the hosting cost being known and acceptable right from the start. I also don't mean to imply that you are not aware of the licensing costs, just wanted to make it clear. For real businesses, at some point it makes more sense to build their own "private cloud" and at this point these issues become relevant but for most start-ups, it's likely just not going to be an issue.
- pistle 12y ago
  While SQL Server has and will continue to be an excellent technical option, the costs can be a drag. That said, the work various .net-related people have out or are working on have the very real availability or promise to lift that dependency, if not provide portability, with DB choice. See RavenDB (document DB) & biggy (Postgres 1st class). Supporting this for more traditional takes on relational store is that there is increasing experience in the .net world for applying coding patterns that can help projects pivot/abstract out the data tool dependencies. If you have a sensitivity in this area, there are options which, when calculated with the ancillary benefits of applying existing .net experience and the great MS free + cloud tools, would pan out very well for many problem spaces. Obviously, YMMV, but if you are doing a wide spectrum of cloud + mobile, Azure is very compelling and Xamarin can become nearly as large a burden cost-wise. That total burden still can be extremely competitive and the ramp for existing .net shops a pretty clear winner. But being a C# developer has never been so promising - and it wasn't ever that awful to begin with. If anything, the problem of coming from the old MS-dev platform-planning and having to understand the range of options and optimizing for costs around those is the weak spot.

viggity 12y ago

a 150 GB business Azure SQL instance is $225/month, sounds like a steal to me.

ryanjshaw 12y ago

> are you going to stay on that version for the lifespan of the system? Likely not. Not sure why you say that -- if it works, it works. I've just spent the last year working on a multi-million dollar project budget building a new system on a SQL Server 2005 cluster. Why? Because there was no need to budget for new hardware. (SQL Server 2008 R2 would have been a big step up, after that not really worthwhile upgrades, in our case.) > For example, if my Azure costs are $10K a month, from that cost is likely $2K-$3K the Microsoft tax. Keep scaling your VMs and servers and that margin starts to look a bit less palatable. Let's say my hosting costs are 10k p.m, 120k p.a. and represent 15% of revenue (this is just how my business works - if I need that much SQL power I'm doing a hell of a lot of business). 120/0.15 = 800k in revenue. Keeping things really simple, let's say my gross margin is 50% and corp tax is 20%. That leaves the business with a nice 320k profit. Sure I don't like spending 4.5% of my revenue on MS licensing, but at this point I can quit my job and port over to Postgresql (or more likely pay somebody else to). The point is that nobody wants to give away 4.5% but if it's the line between success and failure then you probably have bigger issues with your business model/environment. Don't get me wrong - for my current side project I'd love to use Postgresql as the backend and .NET for the rest but the problem with Postgresql is two-fold and really specific to me: 1. I'd spend a lot of my time learning a new set of tools for zero benefit when time is my most expensive resource right now. I'm very familiar with T-SQL and the MSSQL platform (even simple things like having MERGE available has saved me lots of time given my problem domain, nevermind the ability to reuse a lot of code I've written and optimised over the past 10 years and knowledge of various issues and errors I will encounter). Porting in 3 years will be easier because we'll have a working system to port from, and the risk to the project will be less because we aren't doing two things at once (learn a new platform + solve a new problem -- only the later is going to make me money). 2. I plan on using some of the MS BI tools for the first time as part of this project -- this represents quite valuable experience to have as a .NET developer/solution designer; if I go the Postgresql route I will spend a lot of time exploring and evaluating open source options with zero benefit to my day job career progression over the next 3 years - I don't have the luxury of neglecting my day job while I work on side projects, so I need to be careful I don't lose value

avenger123 12y ago