External reviews are not included in the AWS star rating for the product.
I use TiDB Cloud for my PAT environment to test some features in TiDB.
I have used TiDB Cloud to test TiFlash and TiKV performance and compare that with PostgreSQL, which I am currently working with.
I explored features such as analytical query performance with TiDB Cloud, where I examined the options in distributed systems and how a distributed database can help with data processing and other tasks.
According to my experience, TiDB Cloud is very easy to use because we don't need to maintain much in TiDB. All the activities are mostly automated there, and we can easily monitor the performance. We don't need to manually check the health checks and other statistics.
The automation in TiDB Cloud helps my daily workflow as it can generate automatic charts according to our workloads, and we can analyze the queries easily using this monitoring feature. Unlike traditional databases, we don't need to monitor the performance and other statistics manually. Instead, in TiDB Cloud we can use the features it comes with.
I think TiDB Cloud is very useful for beginners because of one feature I have noticed: Text to SQL. This feature means we don't need to write exact queries.
The distributed database options in TiDB Cloud have helped me because we can place the data in multiple data centers and multiple nodes. Due to this, when we are accessing some data, it won't affect all the nodes. It will only focus on the node where the data is present, and other users can still use the other data centers to fetch the data effectively.
From my point of view, TiDB Cloud requires more nodes to run the environment because for a self-hosted environment it needs to deploy three PD, three TiKV, and two TiFlash minimum. I think this is more costly than our traditional databases such as MySQL, MongoDB, or PostgreSQL.
We can improve the performance of TiDB Cloud for customers by reducing the cost and enhancing features such as processing large data by using minimal resources.
The main concern for my score of eight out of ten is about the cost. Additionally, due to its cluster and distributed architecture, it is not beginner-friendly for freshers and starting customers.
As I already mentioned, the cost is the main concern in TiDB Cloud, and also its architecture needs to be made more user-friendly for beginners.
I have been working with TiDB Cloud for more than one year.
The most important advice I will give to others looking into using TiDB Cloud is that cluster-level knowledge such as distributed systems is necessary. We need to understand the architecture and the flow of data and other details about the cloud. Additionally, we need to know about analytical and transactional workloads, which is the most important thing about TiDB Cloud database because it is designed for that. My review rating for TiDB Cloud is eight out of ten.
As a native TiDB user, I install TiDB on EC2 machines, but I wanted to see how TiDB Cloud helps in seamless integration with that ecosystem, so I conducted an experiment with TiDB with Vector and other features.
I have a use case where I wanted to try Vector Search, so I thought of using TiDB Cloud, and I conducted experiments and figured out the use cases for that.
The best features TiDB Cloud offers are mainly because of the easy scale-up of compute and data nodes. On an on-premises cluster that I manage, I would have to procure a new machine and add that to TiKV, but instead, I can easily manage the TiDB nodes, which is the main advantage that I see with it.
In an existing cluster that I manage on-premises, I have to get a new machine, do patching, and make the user SSH, but with TiDB Cloud, I can simply click and scale independently, which is the main advantage that I see with it.
The main advantage that I see with TiDB Cloud is the easy scale-up and scale-in without the need to buy any machines or do patching. It is less hassle; it used to take at least a couple of hours to bring in a new node, patching, and all those other tasks, but with TiDB Cloud, it takes very few minutes, probably a couple of minutes to bring up a new node.
Definitely, the employee and the team will be saved because they do not need a DBA to exclusively set the servers, patching, and all those other tasks, but for money saved, I am not sure about this point since I have not had a real adoption of that.
I think my use case has a direct migration DM tool, which brings the data from MySQL to TiDB layer, and I am not sure if that is available on TiDB Cloud, so that is one thing I might not be aware of, or else that feature may not be present in TiDB Cloud.
One thing I want for TiDB Cloud is the DM because I specifically pull the data from MySQL and put it on TiDB, but with TiDB Cloud, I do not think that DM is available.
Having said that, the one feature that I am exclusively using is the direct migration from MySQL to TiDB, which is a continuous replication from MySQL to TiDB, and I think it lacks in TiDB Cloud.
If you do not have an option to have a replication from MySQL to TiDB, you can go ahead and simply use TiDB Cloud because it is far more efficient than maintaining TiDB on-premises, but if you have a DM kind of solution, which is not yet there in TiDB Cloud, you might have to stick around with native TiDB.
I have been using TiDB Cloud for six or seven months.
TiDB Cloud is stable in my experience.
Scalability is very positive for TiDB Cloud, as DBAs can easily add a node if they figure out the nodes are not sufficient to handle the traffic, so I can give scalability a rating of ten out of ten.
I have not interacted with customer support for TiDB Cloud yet, as I conducted some experiments and have not fully moved to production, hence I have not tested the customer support.
The initial setup with TiDB Cloud is straightforward; the manuals, documentation, and everything were much simpler and easier to execute, so I have not had a hard time doing that.
Performance-wise, I do not see a problem with TiDB Cloud because I already use TiDB on-premises, which is pretty much similar to TiDB Cloud, so I am not seeing much of a difference.
I am partners with PingCAP.
Since I found that the DM is not there, I have not gone beyond the pricing point.
I heard about the pricing because I conducted only an experiment; since I found that the DM is not there, I have not gone beyond the pricing point.
Everything is simpler to adapt in TiDB Cloud versus a normal TiDB, apart from that, I do not wish for anything else to be easier or more flexible. I would rate this review an eight out of ten.
My main use case for TiDB Cloud is that I have used it for testing the vector search for my internal testing purposes, and I have also checked the high availability and auto scale features in TiDB Cloud.
In my internal testing, I have a scenario where I need to build a small application with vector search, and for that, I have deployed TiDB in the cloud using the vector search feature by feeding some text or content and training the model to fetch data with that vector search index.
Regarding my main use case, I find that although it is somewhat costly, TiDB Cloud is best for high availability use cases because we do not need to add the number of PD nodes or TiFlash nodes; it scales up according to our workload, and we do not need to worry about server crashes or resource over-utilization problems as it automatically scales up and down based on our workloads.
TiDB Cloud is best for high availability use cases, and we can also explore more AI features including vector search.
I would like to add that we can enhance security using TiDB Cloud with private and public endpoints, ensuring that our data remains secure, especially for high-security use cases involving very confidential data, making this feature one of the best.
TiDB Cloud has positively impacted my organization because using the vector search feature has made me explore more about it in TiDB, and I have tried to implement that in our organization, and I am currently working on it.
I have explored a feature such as text-to-query, which allows us to avoid writing raw queries; we can just input our own sentence, and the text-to-query feature converts the text into a query which can then be used in the database.
While TiDB Cloud is good for high availability, I think there are some bugs when using the TiFlash feature, sometimes in the TiKV component as well. In TiFlash, I have discovered that it is not suitable for high concurrency since increasing the query concurrency causes TiFlash to overload, sometimes resulting in query failures, while TiKV handles concurrency better but is more focused on row-based data for smaller amounts of data. It would be great if these issues in TiFlash with high concurrency are effectively managed.
According to user experience, I find TiDB Cloud to be more useful for bigger queries, but I also notice complications occurring at times.
When going through the logs, I sometimes find that there are no details regarding errors. I have noticed that while the TiDB documentation lists some error codes to troubleshoot issues, it lacks workarounds or solutions for those error codes, making it not very customer-friendly. I experienced this when I faced an error and could find the error code but not the solution.
I have been using TiDB Cloud for more than one year.
TiDB Cloud is stable based on our deployments. If it is deployed in a cross-region or uses placement rules, it remains stable due to data storage in multiple availability zones.
The scalability of TiDB Cloud is automatically configured, allowing it to scale out when workloads increase and automatically scale in when workloads decrease, which means we do not need to worry.
When discussing customer support, I find that the TiDB community is very active compared to other databases. They quickly respond to questions or feature requests, and when I submitted bug reports, the community provided solutions within three business days, which I appreciate.
Previously, I have used self-deployed TiDB, but I find TiDB Cloud to be better than the self-deployed clusters.
I previously explored TiDB in self-deployment, which led me to evaluate TiDB Cloud as well before starting with it.
I see a return on investment with TiDB Cloud as we reduce the time spent on maintaining and monitoring since everything is automated, including the auto-scaling feature when large workloads arise. Monitoring is deployed automatically, relieving us from manual monitoring responsibilities, which is a significant part concerning ROI.
I would rate TiDB Cloud 10 out of 8; the two points I reduced are due to cost concerns and issues related to user-friendliness.
I take off those two points mainly because of the cost and the user-friendliness issues experienced.
My advice to others considering TiDB Cloud is to keep the cost in mind. If cost is not a concern, I believe TiDB Cloud is the best option among traditional databases due to its distributed architecture which supports high availability and auto-scaling. I also think the TiFlash engine is a good choice for analytical data loads, so I strongly suggest going with TiDB Cloud.
The only additional thought I have is regarding the cost since many customers are worried about it.
I think the questions asked for the reviews are good, but it would help if you waited longer for answers, as sometimes you move to the next question too quickly.
My overall rating for TiDB Cloud is 8 out of 10.