On 26/01/2017 5:38 a.m., erdosain9 wrote: > Hi,........ no > > [root@squid ~]# df -h > S.ficheros Tamaño Usados Disp Uso% Montado en > /dev/mapper/centos-root 48G 16G 33G 32% / > devtmpfs 896M 0 896M 0% /dev > tmpfs 906M 2,1M 904M 1% /dev/shm > tmpfs 906M 8,5M 898M 1% /run > tmpfs 906M 0 906M 0% /sys/fs/cgroup > /dev/sda1 497M 141M 356M 29% /boot > tmpfs 182M 0 182M 0% /run/user/0 > > > by the way, this error dosent appear anymore, but was the first error i > noticed after the bad reboot. (i think that maybe i fix that with "squid > -z".... I think you changed the size of a cache_dir right? The log messages seems to be Squid busily trying to clear the directory size down to what is configured, while systemd keeps trying to force start and restart of overlapping processes mid-action. > > some other approach?? > Not using systemd to control Squid-3. The two are not compatible. As you just found out the hard way. Squid is not a daemon, it is a Daemon + Manager in one binary/process. systemd is based around the naive assumption that everything is a simple daemon and gets horribly confuzled when reality bites. It is not alone, upstart has the same issues. Basically only start/stop work, and even those only most of the time if done very carefully. Your choices with systemd are (1) use the 'squid -k' commands, or (2) upgrade to Squid-4 and install the tools/systemd/squid.service file we provide for that version. Amos _______________________________________________ squid-users mailing list squid-users@xxxxxxxxxxxxxxxxxxxxx http://lists.squid-cache.org/listinfo/squid-users
