Filesystem checkpointing/snapshotting is one of the nice things about ZFS.
If the OS supports ZFS on boot drives, you can do checkpointed full system/kernel upgrades.
I like your idea about process checkpointing. Of course any side effects (eventually we're only talking about network in/out) would be nonreversible, but would be theoretically replayable (have to checkpoint the RTC too). The other side of the connection might have other ideas...
If the OS supports ZFS on boot drives, you can do checkpointed full system/kernel upgrades.
I like your idea about process checkpointing. Of course any side effects (eventually we're only talking about network in/out) would be nonreversible, but would be theoretically replayable (have to checkpoint the RTC too). The other side of the connection might have other ideas...
Sounds like a dream debugging tool though.