The HPE ML Data Management Shell is a special-purpose shell for HPE ML Data Management that provides
auto-suggesting as you type. New HPE ML Data Management users will find this user-friendly
shell especially appealing as it helps to learn pachctl, type commands
faster, and displays useful information about the objects you are interacting
with. This new shell does not supersede the classic use of pachctl shell
in your standard terminal, but is a compelling convenience for power users
and beginners alike. If you prefer to use just pachctl, you can continue to
do so.
To enter the HPE ML Data Management Shell, type:
pachctl shell
When you enter pachctl shell, your prompt changes to display your current
HPE ML Data Management context, as well as displays a list of available commands in a
drop-down list.
To scroll through the list, press TAB and then use arrows to move up or
down. Press SPACE to select a command.
When in the HPE ML Data Management Shell, you do not need to prepend your commands with
pachctl because HPE ML Data Management does that for you automatically behind the
scenes. For example, instead of running pachctl list repo, run list repo:
With nested commands, pachctl shell can do even more. For example, if you
type list file <repo>@<branch>/, you can preview and select files from that
branch:
Similarly, you can select a commit:
Exit the HPE ML Data Management Shell #
To exit the HPE ML Data Management Shell, press CTRL-D or type exit.
Clearing Cached Completions #
To optimize performance and achieve faster response time, the HPE ML Data Management Shell caches completion results. You can clear this cache by pressing F5 forcing the HPE ML Data Management Shell to send requests to the server for new completions.
Clearing the screen #
To clear the screen, press CTRL-L.
Limitations #
The HPE ML Data Management Shell does not support standard UNIX commands or kubectl commands.
To run them, exit the HPE ML Data Management Shell or run the commands in a different terminal
window.