H3C S5600 SERIES Operation Manual page 979

Patches can be added to a patch file incrementally. That is, any subsequent patch file contains all the
patches for fixing errors in the previous patch file besides the patches for fixing the current errors. In this
way, all the system errors found can be fixed once and for all by loading the latest patch file.
In the device, a patch can be in one of the following four states:
IDLE: The patch is initialized but not loaded.
DEACTIVE: The patch is loaded but not activated.
ACTIVE: The patch is active but not running.
RUNNING: The patch is running.
The difference between active patches and running patches:
After system reboot, an ACTIVE patch becomes DEACTIVE and stops working;
After system reboot, a running patch remains running.
The active state provides a buffer to help prevent the system from running with errors introduced by the
by the patches.
Any state of the four cannot be switched to another without user intervention on the command line.
Figure 4-1 illustrates how patch states change with commands:
Figure 4-1 Relationship between patch state transition and commands
To dynamically upgrade a fabric constituted by several switches, you can activate, run, and delete the
patches on the fabric, which can be regarded as an independent device, instead of performing
upgrades on each device. However, for each unit device, patch files need to be uploaded separately
through FTP or TFTP to the flash memory and loaded into the memory with the patch load command.
Follow these steps to perform hot patch configuration:
To do...
Enter system view
Load a patch file
Activate patches
Run patches
Use the command...
system-view
patch load filename
patch activate
patch run
4-4
Remarks
—
Required
Required
Optional