Batching RTK get query

[nouman91]

Hello, and apologies if this has been asked or answered elsewhere—I tried searching but couldn't find relevant information.

I have a use case where I need to fetch data for hundreds of UUIDs. However, our Nginx configuration does not support large query parameters, so I am considering batching the requests. I wanted to check if RTK Query has built-in support for batch fetching or if there's a recommended approach for this scenario.

My use case:

graphQuery: () => {
  querFn: () => {
    const graph = fetchBaseGraph();
    const nodes1UUIDs = graph.nodes.map(node => node.node1);
    const nodes2UUIDs = graph.nodes.map(node => node.node2);
    const nodes3UUIDs = graph.nodes.map(node => node.node3);

    /** Each type of node corresponds to a different endpoint */
    fetchNode1WithBatch(nodes1UUIDs);
    fetchNode2WithBatch(nodes2UUIDs);
    fetchNode3WithBatch(nodes3UUIDs);
  };
};

Each node type requires its own endpoint, and I need to process these requests in batches due to the query length limitation. Does RTK Query offer support for batching, or is there a recommended pattern for handling this kind of use case?

Thank you for your time and guidance!

[markerikson]

There's nothing directly built in for that. Part of it is that RTKQ is not a "normalized" caching tool:

• https://redux-toolkit.js.org/rtk-query/usage/cache-behavior#no-normalized-or-de-duplicated-cache

but also, it's ultimately about tracking the status and contents of a single promise per cache key.

Based on this example, are you intended for each of these items to end up as a separate cache entry? Or do these many items constitute a single cache entry, combined together?

If you're trying to treat these as separate cache entries in the end, we did recently add a new upsertQueryEntries API that's intended for efficiently inserting many cache entries at once:

• https://redux-toolkit.js.org/rtk-query/api/created-api/api-slice-utils#upsertqueryentries

Looking at your example, you could do the actual fetching work yourself (and treat this graphQuery endpoint as a sort of dummy, where you don't care about the actual saved cache entry but rather are using it to initiate the request), then use upsertQueryEntries to actually fill in the individual item cache entries.