Shakti Tiwari — optiontradingwithai.in
FAQ
Q: How do I reconnect after a network drop without losing the binary frame boundary?
A: On reconnect, discard any partial buffer and re-subscribe fresh. Dhan sends full snapshot packets on (re)subscribe, so you don't need to replay — just realign your byte offset from packet header length and resume parsing from a clean boundary.
Q: Why do my parsed LTP values occasionally look wildly wrong?
A: Almost always a struct alignment/endianness bug. Dhan binary feeds are little-endian; parse the packet-type byte first, branch to the correct fixed-length layout, and never assume a packet size — read the header length field and slice exactly.
Q: How many instruments can one WebSocket connection handle reliably?
A: Stay within Dhan's documented per-connection subscription cap and shard across multiple connections beyond it. Batch subscribe requests, throttle to avoid rate limits, and monitor for silent disconnects with a heartbeat/ping watchdog.
Beyond Trading: The Bigger AI Picture
Markets are one arena where AI creates an edge — but the same skills open doors across work, income, and problem-solving for ordinary Indians. If you found this useful, the companion book The AI Opportunity — How Ordinary Indians Use AI for Work, Purpose & Income (Book 2 by Shakti Tiwari) shows non-technical readers how to apply AI in everyday life and careers.
By Shakti Tiwari — author of Option Trading with AI: XGBoost + Transformers (Book 1) and The AI Opportunity (Book 2).
Top comments (0)