Contact Center Video Conference demo throwing connection error via streamingRecognize callback

AI & Data Science Deep Learning (Training & Inference) Riva
1

Trying to run the sample app at Sample Apps — NVIDIA Riva and getting the following error. Any ideas as to what the issue is? I’m running on AWS EC2 and the security group is allowing all inbound traffic to/from my IP. Otherwise I followed the instructions verbatim (except needing to modify the link to the image from “nvcr.io/nvidia/jarvis-speech-client:1.2.1-beta-samples” to “nvcr.io/nvidia/jarvis/jarvis-speech-client:1.2.1-beta-samples”)

Running server on port 8009
Client connected from ::ffff:[my ip]
Initializing Jarvis ASR
Error via streamingRecognize callback
Error: 14 UNAVAILABLE: failed to connect to all addresses
at Object.exports.createStatusError (/workspace/samples/jarvis-contact/node_modules/grpc/src/common.js:91:15)
at ClientDuplexStream._emitStatusIfDone (/workspace/samples/jarvis-contact/node_modules/grpc/src/client.js:233:26)
at ClientDuplexStream._receiveStatus (/workspace/samples/jarvis-contact/node_modules/grpc/src/client.js:211:8)
at Object.onReceiveStatus (/workspace/samples/jarvis-contact/node_modules/grpc/src/client_interceptors.js:1311:15)
at InterceptingListener._callNext (/workspace/samples/jarvis-contact/node_modules/grpc/src/client_interceptors.js:568:42)
at InterceptingListener.onReceiveStatus (/workspace/samples/jarvis-contact/node_modules/grpc/src/client_interceptors.js:618:8)
at /workspace/samples/jarvis-contact/node_modules/grpc/src/client_interceptors.js:1127:18 {
code: 14,
metadata: Metadata { _internal_repr: {}, flags: 0 },
details: ‘failed to connect to all addresses’
}

2

Hi @brian.b
Could you please share the log files and system info so we can help better?
docker logs jarvis-speech

Thanks

3

log_jarvis-speech.txt (53.2 KB)
log_node-js-server.txt (1.1 KB)

Attached. Thanks in advance for the help!

4

@SunilJB , @brian.b ,

I am facing the same issue with updated riva-speech. Any update?

5

Unfortunately no. No one from NVIDIA ever got back to me after I posted the logs and I was unable to figure it out on my own. It’s too bad, because it’s a very interesting use case that I would have liked to seen live.

7

Hi @brian.b
Apologies for delayed response. We will look into this on priority and update you.

Thanks

8

Hi @brian.b
Just wanted to confirm, does the security group has port forward enabled for ports 8009 and 9000?
Also, does config file env.txt has updated value set for URL of your running Riva AI Services?

Thanks

9

Hi @SunilJB and @SunilJB ,

Thanks for your reply. I confirm that in my case, the issue is resolved by changing IP address in env.txt file.

# Configuration environment vars for Riva Contact

# Replace the IP and port with your hosted RIVA endpoint
#RIVA_API_URL="0.0.0.0:50051"
RIVA_API_URL="YOUR_SERVER_IP_HERE:50051" # <------ CHANGE HERE!!!
# NER model to use. This one is the default from the Riva Quick Start setup
RIVA_NER_MODEL="riva_ner"
# NER entities to use from the above model (can be a subset of what is offered)
RIVA_NER_ENTITIES="per,loc,org,time,misc"

# The port your Node.js app will be hosted at
PORT="8009"
# Port for the peer-js server, to be used for negotiating the peer-to-peer chat connection
PEERJS_PORT="9000"
1 Like