A newbie question. I’ve just connected to an AWS hosted Omnisci db (AMI GPU boys version 5.6.4), following the instructions. I can login to Immerse fine. When I connect through ssh, there are no files and no omnisci_path to activate omnisql. What am I missing?
We are sorry, that we aren’t giving clear guidelines for advanced users like you, but on AWS, we are using Docker’s images, so to access tools like omnisql and others, you have to connect to Docker that’s running the omniscidb
So, after connecting
to the host with the centos account, you have to connect to the docker with this command
sudo docker exec -it omniscidb /bin/bash
and then to run omnisql or other commands, you have to type
bin/omnisql [-p Instance-ID]
Thanks for pointing this problem to us and I want so say that I’m truly sorry for the later reply; if you have further questions or remarks, feel free to post in this thread
As a further question, if I wanted to make my own AMI, for the purposes of easily having a persistent data volume, is there a docker-compose.yml etc I can crib from? The instructions for Jupiter install (JupyterLab Installation and Configuration - OmniSci Docs) seemed similar but the AWS AMI loads a separate web server container as well as the database.
asked internally if we can supply the docker-compose.yml we use for our AMI.
I find-out a docker-compose.yml file in the omnisci_data directory located in /var/lib/omnisci/omnisci-storage, and it should be the same file used to create the dockers, the AMI is using.
docker-compose.zip (2,0 KB)
This should create 3 separate containers, making you able to change the configuration of the web part without having to restart the database.