...
When ecFlow needs to write a checkpoint file it first moves (renames) the previous file ECF_CHECK to ECF_CHECKOLD and then creates a new file with the name ECF_CHECK. This means that you should always have a file that is good. If a crash happens while writing ECF_CHECK, you can still recover from ECF_CHECKOLD (by copying its contents to ECF_CHECK), although that version is not quite as up to date.
| Info |
|---|
Tip: You can copy the checkpoint files between systems. |
Another ecFlow server can be started with the original server's checkpoint file and take over from the original ecFlow server host in case of a catastrophic system failure.