Axios 1.5.0 Alters Default Timeout, Impacting Large Dataset Pagination

Direct Answers

maxRedirects: Set maxRedirects: 0 when pagination links arrive via 302 redirects and you need to extract the Location header yourself. The default of 5 will follow redirects automatically, discarding the intermediate headers you need for pagination.

Streaming backpressure: Use responseType: 'stream' with stream.pipeline() (Node 10+) and a modest highWaterMark (e.g., 64 KB). pipeline handles backpressure automatically; manual pause/resume is only needed for custom consumers.

validateStatus + transformResponse: Yes. A custom validateStatus that returns true for 302 and 429 lets the promise resolve so your transformResponse chain runs. The default timeout: 0 (no timeout) is unaffected by these options.

Confirmed Behavior (Axios 1.5.0)

• timeout defaults to 0 (no automatic abort). You must set an explicit value (e.g., timeout: 15000) if you want request cancellation.
• maxRedirects defaults to 5. Setting it to 0 disables automatic following; you must read response.headers.location and issue the next request yourself.
• validateStatus defaults to status >= 200 && status < 300. Override to status => [200, 201, 302, 429].includes(status) (or similar) to treat redirects and rate‑limit responses as successful resolutions.
• transformResponse functions execute in array order. Place header extraction first, then data transformation.

Minimal Working Pattern

const axios = require('axios');
const { pipeline } = require('stream/promises');
const fs = require('fs');

const client = axios.create({
  timeout: 15000,                 // explicit timeout
  maxRedirects: 0,                // handle 302 manually
  validateStatus: s => [200, 201, 302, 429].includes(s),
  responseType: 'stream',         // for large file downloads
  transformResponse: [
    (data, headers) => {
      // 1st: capture pagination headers before data is consumed
      const nextLink = headers.link || headers['x-next-page'];
      const total = headers['x-total-count'];
      return { data, headers: { nextLink, total } };
    },
    ({ data, headers }) => {
      // 2nd: return only what the caller needs
      return { stream: data, pagination: headers };
    }
  ]
});

async function fetchPage(url, destPath) {
  const { data: stream, pagination } = await client.get(url);
  await pipeline(stream, fs.createWriteStream(destPath, { highWaterMark: 64 * 1024 }));
  return pagination; // { nextLink, total }
}

One Diagnostic Detail Needed

Does your API return the next‑page URL in a Link header (RFC 5988), a custom header like X‑Next‑Page, or inside the JSON body of the 302 response? The header name determines the exact extraction logic in the first transformResponse function.