• 0 to 100MB: Flat $5.00 / month
• > 100MB to 1GB: Flat $9.99 / month
• > 1GB to 10GB: $9.99 for first GB, $4.00 for each additional GB / month
• > 10GB to 50GB: $45.99 for first 10GB, $2.00 each additional GB / month
• > 50GB to 150GB: $125.99 for first 50GB, $1.00 for each additional GB / month
To bring that home, let's show what that looks like for some select database sizes:
Size | Cost/Month | Decrease Over Previous Pricing | Effective Cost/GB |
---|---|---|---|
100MB | $5.00 | N/A | |
1GB | $9.99 | 0% | $9.99 |
5GB | $25.99 | 48% | $5.20 |
10GB | $45.99 | 54% | $4.60 |
25GB | $75.99 | 75% | $3.04 |
50GB | $125.99 | 75% | $2.52 |
100GB | $175.99 | 65% | $1.76 |
150GB | $225.99 | 55% | $1.51 |
The calculator on Azure.com has already been updated to reflect the new pricing.
Since most of the world's databases are in fact very small, the new $5 for up to 100MB option is extremely attractive. There's also good news here for those with larger databases. Historically, SQL Azure Database has been one of the more costly parts of the platform. If your database needs were on the larger side (50GB or larger) you were looking at $500/month per database, and if you wanted additional SQL Azure databases for backups or failover that only multiplied the expense. That put many companies in the position of having to choose between the value of having a relational database vs. the cost savings of using alternative forms of storage such as Windows Azure Table Storage which previously was 66 times less expensive. Although NoSQL approaches are becoming increasingly popular, operational cost alone shouldn't be the driver for deciding whether to use a relational database or not. While I've seen some development teams at some companies switch over to table storage easily, for some it's a tough change and they never quite recover the velocity they had with SQL Server. In other words, operational cost isn't the only factor and development team impact should figure stronly into such decisions. With this greatly decreased SQL Azure pricing, and the new low-cost option for 100MB databases, the to-database-or-not-to-database decision can be made on more solid grounds.
No comments:
Post a Comment