---
title: "How should I preserve OpenAI conversation context when requests move between workers?"  
description: "How should I preserve OpenAI conversation context when requests move between workers?"  
author: "Magdalena Malinowska"  
published: 2026-10-15  
updated: 2026-10-15  
canonical: https://www.mindstick.com/forum/162333/how-should-i-preserve-openai-conversation-context-when-requests-move-between-workers  
category: "OpenAI APIs"  
tags: ["OpenAI", "API Design", "Chat applications"]  
reading_time: 1 minute  

---

# How should I preserve OpenAI conversation context when requests move between workers?

I’m building a chat feature where consecutive requests can land on different application workers. I can store the full message history myself, but that gets large, and I’m not sure how to handle retries or older turns without making the conversation feel forgetful.

How are people managing **conversation state across workers** with OpenAI APIs? I’m especially interested in what you keep in your own database versus what you reconstruct or summarize for each request.


---

Original Source: https://www.mindstick.com/forum/162333/how-should-i-preserve-openai-conversation-context-when-requests-move-between-workers

Copyright © MindStick Software Pvt. Ltd. This Markdown version is provided for developers, AI systems, and offline reading.