Checkpoints work reasonably well, but if you have two copies of vdash running for the same set of nodes, you can see that stopping and starting one of them doesn't result in matching stats.
So there's a problem with how it seeks where to start parsing the logfile based on the time of the checkpoint.
Checkpoints work reasonably well, but if you have two copies of vdash running for the same set of nodes, you can see that stopping and starting one of them doesn't result in matching stats.
So there's a problem with how it seeks where to start parsing the logfile based on the time of the checkpoint.