MariaDB Cloud (formerly SkySQL)

SkySQL serves as our transactional layer where we store metadata for all our applications before moving data into our Snowflake database. We receive transactional data from various applications, store it first in SkySQL, then transform it, load it into an S3 bucket, and finally into Snowflake for reporting purposes.

We chose SkySQL because Snowflake is used only for reporting and does not have a transactional layer. Previously, we used a SQL Server database before migrating to SkySQL. We load transactional data first because SkySQL is fully managed in the cloud, eliminating concerns about scalability or performance. We often receive high volumes of transactions from applications, which made SkySQL an ideal choice. After loading all transactional data into SkySQL, we convert that data into aggregate formats for further reporting.

Since we do not have to manage servers and memory as they are completely managed by the cloud, this has been extremely beneficial. Whenever we receive high volumes of transactions, the database auto-scales and improves performance automatically. This is particularly helpful because we experience variable transaction loads, and the auto-scaling and down-scaling capabilities provide cost optimization benefits.

As a data engineer, the most valuable aspect is not having to worry about scalability or performance. I can focus on the actual data itself: whether it is dirty or clean, and what transformations I need to apply. I work on the actual data rather than on infrastructure tasks related to servers.

All our applications are integrated with SkySQL only, and the integration capabilities are powerful. SkySQL can be integrated with a wide range of databases, cloud providers, and supports multiple frameworks, including NCP. Regarding security, it is highly secure because we conduct penetration testing before pushing data from applications to SkySQL.

Currently, we have not identified significant improvement areas for SkySQL. However, enhanced AI functionality would be beneficial. We use AI in Snowflake but not in SkySQL. If SkySQL could write SQL queries automatically using AI or improve SQL query performance, this would be valuable for future use cases.

When comparing SkySQL to other databases such as Databricks and Snowflake, these platforms offer more AI-related functionality and data lineage capabilities, allowing visibility of the complete data journey from source to target. SkySQL currently lacks data lineage functionality. Since we transform data using stored procedures and views, implementing data lineage functionality would help us track the complete journey of data from source through all intermediate transformation steps to the final layer. I rate SkySQL an eight out of ten.

SkySQL has been stable in my experience with no downtime or issues.

SkySQL's scalability is impressive because it auto-scales automatically. Whenever we receive high volumes of transactions from applications, it auto-scales memory and server utilization without any manual intervention. When transaction loads decrease, the server and memory down-scale automatically. This high scalability means that when we onboard new applications that load data into SkySQL, we do not have to worry about scalability as it is completely handled by the cloud provider.

As a data engineer, I have not personally reached out to customer support. However, during the implementation phase, our infrastructure team contacted their customer support for initial setup and received comprehensive support and full responses from them. I would rate customer support as nine out of ten.

Before SkySQL, we used a SQL Server database. With that database, we required a permanent infrastructure team because whenever we faced performance or memory allocation issues, we had to reach out to the infrastructure team and wait for their availability to address server-related issues. After adopting SkySQL, this process is completely automated. We never face performance issues, and backup and scalability are all managed automatically, which has significantly improved our SLA.

SkySQL was purchased through the AWS Marketplace and is completely managed by our infrastructure team.

We have seen a clear return on investment through team size reduction. Previously, we had a four-member team where two members managed the infrastructure side, handling server management, performance, and server optimization, while two people worked as data engineers on the actual data. After adopting SkySQL, the two infrastructure team members were no longer needed because the platform is managed by the cloud. Currently, we have only a two-member team of data engineers working on actual data. We have reduced our team size by fifty percent.

My experience with pricing, setup cost, and licensing is that it is a one-time cost that provides great flexibility after setup. However, in our project, this was completely managed by our infrastructure team, and we as data engineers were not involved with the licensing or infrastructure aspects.

Before choosing SkySQL, we evaluated other options, including the MySQL database. However, MySQL presented the same challenges requiring a separate infrastructure team. We adopted SkySQL because it functions like MySQL but is completely managed in the cloud.

Reducing our team size has significantly affected efficiency and costs for our organization. This cost efficiency benefit resulted from replacing two infrastructure team members since the infrastructure work is now automated, eliminating the need for a dedicated infrastructure team. This is very helpful for cost optimization and provides complete monitoring capabilities. You can view server health and optimization from a central point, allowing visibility into when you receive higher data volumes and when volumes are lower.

Regarding SkySQL's AI capabilities, we are using the AI functionality to write SQL queries. Whenever we face issues with SQL queries that join multiple tables, we use the AI capability to fix errors automatically. This saves considerable time for data engineers.

The accuracy and reliability of SkySQL's AI output is quite consistent. We conducted a proof of concept comparing manually written SQL queries with queries generated by SkySQL's AI capabilities, and both approaches yielded the same results. This demonstrated that the AI functionality is highly accurate.

My advice to others evaluating SkySQL is that if you are primarily working on analytics, business intelligence, or reporting, other options such as Databricks or Snowflake may provide higher functionality related to analytics and GenAI. However, if you are working with a transactional database or seeking a platform to store transactional data, SkySQL is worth evaluating. I have rated this product an eight out of ten overall.

My main use case for SkySQL is to store and manage asset and site data for an industrial asset performance management application with support for analytics, reporting, and high availability.

SkySQL supports my asset performance management application by helping us build a solution that stores asset maintenance and operational data from multiple manufacturing sites. We need a managed, scalable, and highly available database to support dashboards, analytics, and day-to-day application transactions, and that is why we are evaluating SkySQL.

Our main use case is covered, and we are currently evaluating SkySQL to see whether it meets our scalability, availability, and operational requirements.

In my experience, the best feature that SkySQL offers is a fully managed database service which reduces operational overhead, along with its high availability and automatic failover for business-critical applications, and its scalability to handle growing data volumes and overloads, plus automated backups and disaster recovery capabilities. These are the key capabilities that make SkySQL attractive for an asset performance management application, but we are still validating them as part of our evaluation.

Since we are still evaluating SkySQL, we have already seen an impact, but as we are still in the evaluation phase, I cannot say any feature has had the biggest impact in production yet. The feature that is most important to us is high availability with automatic failover because our asset performance management application supports critical manufacturing operations.

Besides high availability and scalability, we are hoping to see simple administration and reliable performance. Also, since it is a managed service, we would like to minimize database maintenance, patching, and backup management so our team can focus more on developing the asset performance management application.

It is still early in our evaluation of SkySQL, so we do not have major concerns, but if anything, we would like to see even better documentation, onboarding guidance, and monitoring capabilities to make adoption and ongoing management easier.

During evaluation, having more end-to-end onboarding guides with real-world deployment examples would make it easier to validate different architectures. For monitoring, it would be helpful to have more out-of-the-box dashboards, proactive health alerts, and clear recommendations for performance tuning and capacity planning. Those improvements would help teams evaluate and operate the platform more efficiently.

I give SkySQL an eight out of ten instead of a nine or a seven because we are still in the evaluation phase, so I would like to gain more hands-on experience with deployment monitoring, performance under real workloads, and overall operational management before making a final judgment. We also see opportunities for enhanced onboarding guidance and more comprehensive out-of-the-box monitoring and performance recommendations.

We are considering Microsoft Azure and AWS before making a final decision about SkySQL, and we have not finalized the cloud provider yet, but we are considering Microsoft Azure and AWS. It is between the two, and I am not sure which one we are going to deploy.

I have been using SkySQL for a year now.

SkySQL is stable in my experience.

So far, SkySQL appears to scale well and aligns with our future growth requirements. We still need to validate it under real workloads, but the initial evaluation has been encouraging.

We are still in the evaluation phase of SkySQL, so we have not had enough interaction with the support team to evaluate it yet, so it is too early for us to comment on the quality of customer support.

We are comparing SkySQL with other managed database options, and we have not migrated from a specific solution yet. This is part of our evaluation process to identify the best fit for our application.

We have not purchased SkySQL yet as we are still in the evaluation phase. Therefore, no purchase decision has been made, including whether we would use the AWS Marketplace or not.

Our company does not have any business relationship with the vendor other than evaluating SkySQL as a potential customer.

For this question, I would say not yet because we are still evaluating SkySQL, so we do not have quantified ROI metrics. Our expectation is that it will reduce operational overhead and database administration time, but we will need production experience before we can measure those benefits.

My experience with pricing and setup cost regarding SkySQL is that it is still too early to provide a definitive assessment. We are currently reviewing the pricing, licensing, and overall cost of ownership as part of our evaluation.

We are looking at options such as Amazon RDS for MySQL, Azure Database for MySQL, and other managed relational database services before choosing SkySQL. Our goal is to find the solution that best meets our requirements for scalability, reliability, ease of management, and cost.

My advice for others looking into using SkySQL would be to run a proof of concept with your own workloads and validate scalability, availability, ease of management, and cost. Comparing it with other managed database options will help you determine whether it is the right fit for your environment.

Regarding SkySQL, we have discussed it thoroughly, and I do not think there is anything I am able to provide a review about since we are in the testing phase. I give SkySQL an overall rating of eight out of ten.