Initial Setup¶
After installation of P-AIRCARS, before running the pipeline, some initial setup is needed. These include downloading some required metadata for the pipeline. P-AIRCARS also provides multiple ways to monitor its progress – local GUI logger (for local machine only), over email, web-based remote logger, and prefect dashboard.
Download P-AIRCARS metadata¶
To download and save the required P-AIRCARS metadata in appropriate directory, run from command line:
init-paircars-setup --initClick here to see parameters
P-AIRCARS data directory and docker environment is not setup yet usage: init-paircars-setup [-h] [--init] [--port PORT] [--no_kill_port] [--datadir DATADIR] [--update] [--remotelink LINK] [--remote_password PASSWORD] [--emails EMAILS] Initiate P-AIRCARS data options: -h, --help show this help message and exit --init Initiate data --port PORT Prefect port (default: 4260) --no_kill_port Do not kill occupied port --datadir DATADIR User provided data directory (default: ) --update Update existing data --remotelink LINK Set remote log link (default: None) --remote_password PASSWORD Set remote log password (default: None) --emails EMAILS Email addresses (comma seperated) to send Job ID and password for remote logger (default: None)If data files are present, but needs to updated, run:
init-paircars-setup --init --update
Important
prefect server will be automatically setup during this process. In local environment, if there is any issue in starting prefect server, prefect will automatically fall back to its ephemeral mode. For cluster environment, P-AIRCARS can not be run without prefect server. By default, prefect server uses the port 4260 and its associated PostgreSQL database used port 5260. If these port are pre-occupied, P-AIRCARS will try to kill those port to avoid port overload. If not successful, the closest free port will be used. The instructions to access prefect dashboard will be displayed in the terminal with the steps to access it. If user do not want to close the port, use --no_kill_port parameter during initiation.
Custom P-AIRCARS metadata directory¶
By default, data directory will be at “~/.paircarspipe/paircarspipe_data”. It requires 20 GB space. Sometimes home directory may not have sufficient space. In that case, one can setup data directory in a custom location as follows:
init-paircars-setup --init --datadir </full/path/to/custom/datadir>
Note
This is not data directory for observations. We strongly suggest not to change anything in this directory after P-AIRCARS initiation.
Setup e-mail ids¶
To receive remote logger Job ID and password, and e-mail notifications for pipeline progress, use can setup their e-mail id(s) in P-AIRCARS.
init-paircars-setup --init --emails <youremail1@email1.id1>,<youremail2@email2.id2>
If you setup a remote logger as described below, you will receive a unique Job ID and password (user provided or auto-generated six-character) to access logs of a particular pipeline run from the remote logger. Without this password, one can not access logs of that particular pipeline run. This added security as well as privacy when multiple user uses the same remote logger link, for example, an institute based remote logger link.
Setup remote logger link¶
If remote logger is intended to be used, setup the remote link in P-AIRCARS metadata. By default, P-AIRCARS will set an auto-generated 6-character password for you.
init-paircars-setup --init --remotelink https://<remote-logger-name>.onrender.com
Setup custom remote password¶
If user wants to set a custom password for them
init-paircars-setup --init --remotelink https://<remote-logger-name>.onrender.com --remote_password <your-custom-password>
Update remote logger link and/or e-mail ids¶
If user wants to update the already provided remote logger link or e-mail id(s), simply run the above commands with new values. P-AIRCARS will automatically update the database with these new values.
Setup remote logger¶
Before using remote logger this, create your own remote logger on free-tier cloud platform, https://render.com. One can use, same remotelink in multiple machines and users. However, free-tier link has some limitations on bandwidth and concurrency. If you want to use remotelink for your institution, we suggest to purchase suitable paid version or setup seperate remotelink for different users.