gdraheim
systemctl.py does no journal logrotate
When doing "systemctl status cron" then systemd may report
Warning: Journal has been rotated since unit was started. Log output is incomplete or unavailable.
The systemctl.py replacement does create a path_journal_log as well, but it is never rotated. When using a container that runs for a long time then it might get to be a large file.
As the processes of the services are completely independent one can just exchange the out file-node where to write to. An option may be to use "falloate --punch-hole" to put sparse blocks into the journal.
If one does really want to have a rotatation one could do some "dd" first. However I would prefer to not do such file creation - or even do it like just once a day.
Instead I would like to implement an option where the journal is given a maximum non-sparse value, and everything below is getting punched in the init_loop. That should keep the interruptions rather short - just one hole at a time most of the times.
fallocate --punch-hole --offset 40K --length 4K service.log
# TEST
dd if=/dev/random of=random1.log count=200
dd if=/dev/random of=random2.log count=200
fallocate -p -o 0 -l 12K random1.log
du random*
16 random2.log
4 random.log
cat random1.log | od -a | wc -l
223
cat random2.log | od -a | wc -l
985
ls -l random*
-rw-r--r-- 15805 10. Sep 16:02 random1.log
-rw-r--r-- 15735 10. Sep 16:03 random2.log
man fallocate(1)
-p, --punch-hole
Deallocates space (i.e., creates a hole) in the byte range starting at offset and continuing for length bytes. Within the specified
range, partial filesystem blocks are zeroed, and whole filesystem blocks are removed from the file. After a successful call, subsequent
reads from this range will return zeroes. This option may not be specified at the same time as the --zero-range option. Also, when
using this option, --keep-size is implied.
Supported for XFS (since Linux 2.6.38), ext4 (since Linux 3.0), Btrfs (since Linux 3.7) and tmpfs (since Linux 3.5).