AnalyticalTable: are there any plans to migrate to @tanstack/react-table (v8)?

[yauhenikashlei]

Since v7 has been unmaintained for several years, it can cause performance limitations when dealing with large datasets.

It would be helpful to understand whether such a migration is on the roadmap or if there are any recommended approaches for handling large data volumes with AnalyticalTable.

[Lukas742]

Hi @yauhenikashlei

Regarding react-table (@tanstack/table) v8:
Migrating the AnalyticalTable to use v8 is not feasible, as the underlying structure has completely changed. There were efforts planned to introduce a new table that leverages the v8 react-table, but they are now on hold for a while, since our capacity is limited and priorities have changed. Early prototyping of a v8-based table showed some minor improvements, though the prototype only covered basic features with minimal overhead, so a direct comparison to the full AnalyticalTable is not meaningful. You can find more information about this here, and to receive updates if development of the new table continues by subscribing to this PR.

Regarding performance:
While react-table v7 is unmaintained, it is stable. We test the implementations the AnalyticalTable uses on top of it and ensure compatibility with React 19. Migrating to v8 could yield minor improvements, but no significant performance gains - client-side sorting, filtering, and grouping inherently needs to iterate over the full dataset regardless of the library used. For very large datasets, server-side processing is the recommended approach (see below).

For handling large datasets, here are some recommended approaches:

• Ensure stable prop references. Basically everything that touches the table state needs to have a stable reference; props like columns, data, tableHooks, selectedRowIds, but also frequently called handlers. Unstable references are the most common cause of performance issues.
• Server-side operations. If the dataset is too large to load entirely in the browser, you can use reactTableOptions={{ manualSortBy: true, manualFilters: true, manualGroupBy: true }} to delegate sorting, filtering, and grouping to the backend. Note that this means the table state is rebuilt whenever new data arrives. To prevent this you can use the respective react-table option.
• Consider memoizing custom Cell renderers. If you use custom Cell components with expensive computations or deep component trees, wrapping them in React.memo can help reduce re-renders when table state changes while rows stay in the viewport.

Important: The most common reason the AnalyticalTable feels slow is development mode overhead — this is not specific to react-table v7 and equally affects v8 or any virtualized table implementation. React's dev mode (Strict Mode double-rendering, validation checks) combined with Chromium's V8 debugger instrumentation causes significant performance degradation for virtualized components. Please see our FAQ entry on dev mode performance for detailed recommendations.

If you're experiencing a specific performance issue, feel free to share a reproducible example and we can take a closer look.