Flexmonster and Apache Arrow stream loading

Hello Varun,

Thank you for contacting us.

Currently, Flexmonster does not support Apache Arrow out of the box. We suggest using one of the following options for using it with Flexmonster:

• Implement a server-side script, which will return the data from Apache Arrow in JSON/CSV format on a certain URL endpoint. 
• Implement the custom data source API to retrieve data from your data source and process queries from Flexmonster on your server.
• For .NET stack: using Flexmonster Data Server as a DLL allows implementing custom data fetching logic.

Please let us know if other questions arise.

Kind regards,
Nadia

Hello Varun,

Thank you for the response.

We want to explain that closing the browser tab releases the browser’s memory, but it does not affect memory on your server container if your own backend is buffering or caching that data.

If you use only client-side processing (without Flexmonster Data Server), Flexmonster runs entirely in the browser. In this case, the container should only stream the data (CSV or JSON) and then release it. If memory is not being released, the issue is likely related to your own API or streaming implementation. We recommend ensuring that the server does not store the full dataset in memory between requests or cache large responses.

If you are using Flexmonster Data Server inside the container, the memory behavior is different. FDS loads data into indexes stored in RAM to answer queries quickly, and those indexes remain in memory even if the browser tab is closed. Memory is released only when an index is deleted, and the .NET garbage collector reclaims the memory, which may not happen immediately.

To minimize and better control memory usage in this case, you can adjust the FDS configuration. For example, if response caching is not required, you can disable it:

"DataStorageOptions": {
  "CacheSizeLimit": 0
}

You can also disable keeping data during refresh to avoid temporarily duplicating data in memory:

"DataStorageOptions": {
  "KeepDataOnRefresh": false
}

You are welcome to check our documentation for more details:

• CacheSizeLimit
• KeepDataOnRefresh

Please let us know if other questions arise.

Kind regards,
Nadia