Today we are excited to introduce the new Observability Dashboard for Data Stores, providing real-time monitoring and debugging capabilities. This dashboard offers three essential charts to enhance your workflow:
Request Count by API: Breaks down the API request count per minute by API method, e.g. GetAsync() so you can understand which API methods are called the most.
Request Count by Status: Groups the API requests based on response type, e.g. OK, TooManyRequests. You can use this information to identify the top issues in your Data Stores usage.
You can also filter by Data Store Type using the Filter By button at the top of this page to show requests for only Standard or Ordered Data Stores. The View Error Report button will take you to the Error Report page with a “DataStoreService” filter populated, which gives you more detailed information like the Data Store name.
FAQ
What request sources are included?
The dashboard shows requests from Game Servers. Stay tuned for insights from Studio and Open Cloud requests.
How do I handle errors?
For more details on the meaning of each error and possible solutions, please consult the documentation for Data Stores error codes.
For errors related to throttling, we recommend using exponential backoff to retry your requests.
A small number of requests to Data Stores may fail due to “InternalError” and/or “InternalServerError.” We also recommend using exponential backoff to retry these requests. However, if the volume of these errors increases significantly over a short period of time, please file a bug report.
We hope you enjoy this update! Please let us know if you have any questions.
Cool update. Would love to test it, but Roblox is down rn.
Will be interesting to see how many requests aren’t going through successfully. I used to have quite a lot of data loss with my own datastore system, but I have never received another complaint since I switched to ProfileService. 10/10 I never have to worry about data loss ever again!
Edit: I’m amazed by how low the amount of unsuccessful requests is in my game! Thank you for these great analytics.
This seems like a really cool update that will be pretty useful in the future, but I do have a request.
Developers should have a way to create their own custom dashboards, with the ability to add data from in game. This would allow developers to create their own tools for tracking performance and issues in game, and would open up all sorts of possibilities.
this definitely seems like a cool feature despite my low knowledge on the DataStoreService, though i’d love to see some improvements to in-studio modeling - adding tons of new tools like extruding, beveling, et cetera - would definitely be a huge improvement. maybe start working towards that next?
…also maybe fix cylinders and spheres because they are not even at all… [i bring up spheres because they align with cylinders on the same orientation]
Awesome update, thanks for providing these new tools. A long-term project I work on suffers from datastore request nightmares (particularly from misusing the datastore API) and enhancements like these will be able to help us identify where & when issues are more likely to occur.
Hi @Zero_Protocol,
Your feature request for in-studio modeling and bug with cylinders are important. Can you file feature requests/bugs on DevForum so we can route the issues to the appropriate teams?
There as a method to the madness if you’re curious. Evenly spaced points on a square (cylinder) / cube (sphere) are taken and projected down to the radius.
It would be great if the charts could display something like “requests per minute per X players” or even better somehow display the datastore limits at that time on the chart. Because the flat number of requests is hard to use when CCU and number of servers is constantly fluctuating.
I do love how detailed we can filter the “Request Count by API x Status” graph, I’m sure that will come in handy at some point!
Great update! Thanks Roblox team.
It would be more helpful if the page has quota information like Memory Stores Monitoring page. As you can see, the limit varies depending on the number of players, making it difficult to calculate the quota usage.
