I really like the concept of the journal. It makes troubleshooting a hell of a lot easier. However, right now it's extremely slow to view a log the first time on btrfs and a rotational disk (a number of people have complained on the mailing list).
For example, "journalctl -b" takes minutes to display anything after a few days worth of logs have been saved. After that, it's quick until a restart.
Hopefully some of this will be fixed in systemd 195, I see a few journal memory management fixes in the log.
A nice to have would be finer grained control over what gets stored persistently. Right now it looks like just a log level is used, but I might have different log levels (or nothing at all) depending on different filters.


Reply With Quote
