Covalent self deployment cloud images

[venkatBala]

Based on the recent PR on adding support for creating cloud VM images using Packer to facilitate self hosting, following are a few suggestions regarding the images

• Create images for all the supported platforms as shown in the compatibility matrix

• Different version of python i.e. v3.8.13, v3.9 ..

• Installing the python and covalent at the system level will facilitate deployment as a standalone service

• The machine image can be configured such that covalent server and UI starts running as soon as a VM is provisioned using it. This allows one to expose covalent as a systemd service with a well defined unit file. The service can be then easily started, stopped and restarted via systemd commands

• Using systemd also allows one to use systemd sockets that can be configured to listen for specific events on specific ports. Systemd sockets, when activated can be configured to trigger a shell script or any other program. This can allow one to gather traffic to the covalent server as well as allow one to build custom logging/data pipelines

• Docker images of the same can also be configured to facilitate container deployment and execution

• If systemd unit files are used, default covalent configuration settings can be pre-installed/baked into the image so that there is a factory setting for covalent that users can use as a starting point

• To further ease installations, @wjcunningham7 suggested developing .deb and .rpm packages for the covalent server (not the SDK) that can be easily installed by apt-get install or dnf/yum install

@wjcunningham7 @scottwn @santoshkumarradha