Many companies are thinking about moving their on-prem MSSQL database workloads to the cloud. However, many of those same companies are also trying to rein in ballooning cloud costs. Enter the multi-cloud strategy.
A multi-cloud strategy revolves around moving certain workloads to certain cloud platforms for financial or product quality reasons. For example, many companies choose to use Snowflake for their analytics workloads.
I recently had the chance to test out Tessell, a multi-cloud vendor that offers managed SQL Server, PostgreSQL, MySQL, and Oracle databases on multiple clouds, and was pleased with their performance, ease of use, and backup features.
Benefits of going multi-cloud
Let's find out the most common talking points for going multi cloud.
"Hey chatGPT, what are 5 benefits of going multi-cloud?"
- Improved resilience and high availability of services by distributing workloads across multiple cloud providers
- Greater flexibility and agility to deploy workloads on different cloud platforms to meet specific requirements
- Increased vendor choice and competition, leading to better pricing and service quality
- Ability to leverage unique capabilities and services of different cloud providers to address specific needs, such as security, compliance, or data management
- Enhanced performance and optimization by distributing workloads to minimize latency and take advantage of the strengths of each cloud provider
Those are pretty much the standard things that you will hear people say about multi-cloud. I agree with all of them, except the point about improving resilience and availability. In the past, I was a pager-carrying member of a multi-cloud devops team, and many of our problems actually came from differences in environments.
Tessell Multi Cloud platform
Tessell is a good case study for the other benefits.
Tessell scores full points for enabling "greater flexibility and agility". The platform allows you to deploy a workload to a specific cloud platform in a specific data center. For example, let's say you need to run your workload in country X but your current cloud provider doesn't operate there. Multi cloud to the rescue! Tessell's list of data centers is growing and currently includes many of the most important AWS and Azure data centers. I imagine they will add more in the future.
Tessell also affords you "Increased vendor choice and competition." Don't like a recent price spike in DB cost by your cloud provider? Switch over!
The last two points, "Ability to leverage unique capabilities and services of different cloud providers" and "Enhanced performance and optimization by distributing workloads" are where I think multi-cloud really shines. For example, I personally like the Bigquery data warehouse more than Redshift because I think it offers much better performance and scalability. However, I like AWS more in general, and spend most of my time working there. In this example, I could adopt Google Cloud to gain access to Bigquery, or instead, just use the multi-cloud vendor Snowflake. It seems like Tessell is aiming to become a "Snowflake of OLTP workloads".
Here is a screenshot of the Tessell UI. As you can see, it has a lot of the options that you would expect from a managed DB service, such as instance size, software release, etc.
While I haven't had a chance to personally test all of Tessell's features, I found the UI intuitive and was able to easily do what I wanted to.
Also, I had an issue with an account permission and was able to get instant support through their built in chat feature. I love that.
How to Migrate an on-prem SQL Server DB to Tessell
Our newest tool, Albatross, was originally designed to move a SQL Server database to PostgreSQL running the Babelfish extension. However, it can also be used to migrate from one SQL Server database to another.
Check out the video at the beginning of the article to see me demo the tool, and transfer an on-prem MSSQL DB to Tessell in just a few minutes!
Albatross Key Features
The key features of Albatross are:
- Albatross catches up the target SQL Server database database to the source SQL Server database without having to do any manual transfers.
- Once the target has completely caught up, the tool allows you to cut your workload from the original database to the new database with zero downtime.
- The tool can detect and log any differences in query results between the original database and the new database.
How does Albatross work?
- Albatross gives you instructions to export your schema with SQL Server Management studio.
- Albatross then instructs you to point your MSSQL clients at itself and then proxies those connections to the original MSSQL database.
- Next, it automatically catches up the target database with the source using a specialized version of SQLpipe's technology.
- Once the target DB is caught up, it makes a copy of every incoming query before sending it along to the source MSSQL DB.
- It runs each query on the target as well, thereby keeping it up to date with the source database.
- Last, it logs the response from both databases to a CSV file so that you can confirm the database is a valid replacement.
This is a lot of information - if you'd like to learn more about how Albatross works, check out the video!