AI Assistant bundle import failed.

Products

• VMware vDefend Firewall
• VMware vDefend Firewall with Advanced Threat Prevention
• VMware Avi Load Balancer

Issue/Introduction

AI Assistant bundle import failed.

Environment

vDefend SSP Version: 5.2 and later

Avi Operations: 5.2 and later

Cause

One of the main reasons the bundle import failed is because during bundle import llm calls could not be made successfully. Few use cases that could cause bundle import failures:

1. vTIS connectivity issue
2. LLM Key is expired
3. When configured with Broadcom account, enough quota is unavailable.
4. Embedding model used in the bundle is not configured or available.

Resolution

Depending on the cause, resolution can be used accordingly.

1. vTIS connectivity issue - Check for open alarm and follow its KB for remediation

2. LLM Key Expired - Check for open alarm and follow its KB. 

3. LLM Quota Exhausted - Check for open alarm and follow its KB. 

4. If you are using Broadcom account, check existing Broadcom quota using GET API /ai-assistants/broadcom-account-config response, quotas > daily_token_quota and current_daily_token_usage to check daily consumed quota. Wait for next day to reset if less tokens available when bundle import was triggered either manually or automatically. Use /ai-assistants/platform/reconcile API to trigger bundle import again. 

5. Check intelligent assist dialog and intelligent-assist-models pod logs for errors and if error is related to unsupported models. Check your cloud provider project details to make sure if project has support for (default) embedding model from /ai-assistants/llm-support?usage=AI_ASSISTANT_PLATFORM. If you are using custom model, check specific model info if supported in your cloud provider project details.  

    

Pod logs (intelligent-assist-vdefend-dialog-0-####)

• Login to SSP-Installer with sysadmin credentials
• Fetch the pod name using "k get pods -A | grep dialog"
• Copy the pod name that looks like intelligent-assist-vdefend-dialog-0-####
• Check the logs of it using the command "k logs intelligent-assist-vdefend-dialog-0-#### -n nsxi-platform" and grep accordingly.

2026-08-03 09:56:41.401, ERROR: [/stage] [build/dialog_srv.cjs:537317:19] [reqid=1785749621634/17]  message processing error {
  message: "Models2ClientRPCImpl.cpp:159 [alan::Models2ClientRPCImpl::embedTexts(std::string, const std::vector<std::string> &, std::string, std::string, int, bool)::<lambda()>] embedText failed: Exception iterating responses: 404 NOT_FOUND. {'error': {'code': 404, 'message': 'Publisher model `projects/<project>/locations/us/publishers/google/models/gemini-embedding-001` was not found or your project does not have access to it. Ensure you are using a valid model name and that the model is available in the specified region. For more information, see: https://docs.cloud.google.com/gemini-enterprise-agent-platform/resources/locations.', 'status': 'NOT_FOUND'}}\n"

To re-import bundle again use below APIs:   

• To delete bundles other than PACKAGED source:  

  • From UI goto Connection Mode & Bundles > AI Assistant Bundles > Edit > Manage Bundles > Delete bundle. 

  • Using API - GET /ssp/security-content/bundles?bundle_type=AI_ASSISTANT_FIREWALL and match the version you want against ssp_version/bundle_id, take its id, then: 

    DELETE /ssp/security-content/bundles/{id}

• To Import bundle again

  • To reimport already imported bundle even if its successful use API - POST /ai-assistants/platform/reconcile?force_bundle_import=true. 

  • To Auto import new bundle - Either enable auto download or manually import new bundle. If auto download is already enabled, toggle auto download to false and true again to trigger auto download new bundle immediately.

• If none of the steps are working, delete existing bundles as above mentioned. Then de-activate and activate AI Assistant in UI.