{
"$type": "site.standard.document",
"bskyPostRef": {
"cid": "bafyreigtu3nmukry6qt273qq2shxn4opaotr6ttox7pgdjn2edv6bdhsr4",
"uri": "at://did:plc:lk3jfj3zq4k4wxnk474axylu/app.bsky.feed.post/3mmyxm7cdktx2"
},
"path": "/t/vector-store-api-calls-returning-504s-503s-and-generally-being-slow/1381938#post_7",
"publishedAt": "2026-05-29T14:55:46.000Z",
"site": "https://community.openai.com",
"tags": [
"@Con",
"@S_z"
],
"textContent": "Hey @Con, that can definitely be frustrating.\n\nSince you've already tried a few approaches, one workaround that has helped others is treating the update as a migration rather than modifying the existing vector store:\n\n * Create a new vector store for the updated resource set\n * Add files in batches to reduce write pressure during ingestion\n * Wait until ingestion is fully complete\n * Run a small smoke test against the new store\n * Update the assistant to use the new vector store ID\n * Keep the previous vector store temporarily as a rollback option\n * Delete the old vector store only after the new one is confirmed stable\n\n\n\nThis isn't ideal, but it can help avoid issues during large updates or re-indexing operations.\n\nProps to @S_z too\n\nAvinash",
"title": "Vector Store API calls returning 504s, 503s, and generally being slow"
}