Multi-Execution
When you need to distribute the same command to multiple servers (like checking clock synchronization or verifying system versions), Multi-Exec drastically reduces repetitive work.
Launch Multi-Exec
Section titled “Launch Multi-Exec”You can select target servers and launch Multi-Exec in two ways:
- From Active Connected Tabs: When you are in the terminal panel (Normal Mode), provided you have at least two or more active Tabs, press
xto target all currently running tabs and initiate bulk execution. - Directly from Connection List: In the Connection List interface, press
xto initiate bulk execution. (Note: TheVkey multi-selection mode in the connection list is primarily for management tasks like bulk deletion; do not confuse it with bulk execution). PhanTerm will establish secure temporary connections in the background to execute the task.
Three-Stage Execution Panel
Section titled “Three-Stage Execution Panel”Multi-Exec employs a secure, three-stage overlay panel design:
- Command Input Stage: Here, you write the Bash command to be executed (you can press
Ctrl+Sto open the Snippet library for quick insertion). - Execution & Cancellation Stage: Upon pressing Enter, the command is dispatched to all target servers in high-concurrency mode. If you spot a mistake in your command, you can quickly press
Escto abort the dispatch; the system will safely retain any partial results already returned. - Result Aggregation Stage: After execution, the stdout/stderr from all servers are displayed in split panels, which you can quickly navigate using
j/k.
Smart Diff Comparison Mode
Section titled “Smart Diff Comparison Mode”During the result presentation stage, if you need to verify whether all machines returned identical output, you can press d to enable Diff Mode.
In this mode:
- All “consistent lines” (lines that are exactly the same across every machine’s output) are dimmed.
- Any lines that differ on some machines or appear on only one machine are highlighted in color and prefixed with an
!warning.
This makes it possible to spot anomalous configurations or un-updated nodes at a glance among the complex outputs of dozens of servers.