Closed suman-maity closed 6 years ago
Do you have any logs from the server that you can share?
The issue is resolved. The EC2_AVAILABILTY_ZONE property has a typo. I corrected it in my conductor server properties file and the problem was gone. I should have closed the issue by now. Closing it now. Thanks.
Polling for tasks is not working when i am running server in In-Memory.
Hi,
I have a single node set up of conductor back-end such as Dynomite, Redis and Elasticsearch on an Amazon EC2 instance. I have conductor server (jar file) running on another EC2 instance. Both instances are in the same availability zone. Creation of task definition, workflow definition and retrieval of in_progress tasks (/api/tasks/in_progress/) are working. However, polling for tasks is not working, None of the /api/tasks/poll/ and /api/tasks/poll/batch/ is working. They return success without any response body.
I have looked into existing posts on this matter but none of those have helped me yet while those resolved the same issue faced by others. I am not sure if I am missing something in my configurations. Below are the configuration for each of Dynomite, Elastic Search and Conductor server.
Redis Redis is running on port 6379.
Dynomite Config YAML file dyn_o_mite: dyn_listen: 0.0.0.0:8101 data_store: 0 listen: 0.0.0.0:8102 dyn_seed_provider: simple_provider rack: us-east-1b servers:
Elastic Search cluster.name: cosmos-elasticsearch-cluster node.name: cosmos-elasticsearch-node-1 network.host: 0.0.0.0
Conductor server properties db=dynomite
workflow.dynomite.cluster.hosts=172.31.60.217:8102:us-east-1b
workflow.dynomite.cluster.name=dyn_o_mite
workflow.dynomite.connection.maxConnsPerHost=31
workflow.namespace.prefix=conductor
workflow.namespace.queue.prefix=conductor_queues
queues.dynomite.threads=10
queues.dynomite.nonQuorum.port=6379
workflow.elasticsearch.url=172.31.60.217:9300
workflow.elasticsearch.index.name=cosmos-elasticsearch-cluster
EC2_AVAILABILTY_ZONE=us-east-1b