hi, 2 times a day at 22:00 and 22:05 i have the follwing error in the systemlog: Code: Error executing jk_init command: ERROR: jail directory basicshell is not safe Usage: /usr/sbin/jk_init [OPTIONS] Usage: /usr/sbin/jk_init [OPTIONS] -j jaildir sections... -h --help : this help screen -c, --configfile=FILE : specify configfile location -l, --list : list all available sections in the configfile -j, --jail= : specify the jail to use. For backwards compatibility, if no jail is specified, the first argument after the options will be used as jail -v, --verbose : show what is being done -f, --force : force overwriting of existing files -k, --hardlink : use hardlinks if possible I saw a thread with the same issue, but not a writeup of a fix. can you help me? https://forum.howtoforge.com/threads/observations-or-bugs-with-ispconfig-3-3-beta-1.94101/
I have not seen this error on any system yet. However, perhaps the other user who experienced the issue can provide insight into what they did to resolve it.
I asked him a while back, but no answer. it started with the latest update... it is a system which has had a lot of updates over the years.
I suppose that the regular jailkit update job gets an empty path for two jails on your system, which results in the error message. Because basicshell is not a path, its a section. So instead of something like: jk_init -v -j /var/www/clients/client0/web1 basicshell the path is empty, which results in: jk_init -v -j basicshell and this triggers the error message. So, somehow, there might be two websites in the web_domain table (websites are the type vhost) that have no path.
You can check the web_domain table if you have records of type vhost in there without a path. Other records, like subdomains etx, have no path; that's fine, this is only about type vhost records.
Yes, I mean the document root. Then I'll have to check the code to see what might cause an empty path.
It might be worth trying to enable debug log mode under System > Server config, maybe this gives us some insights into which site is affected when the issue occurs next time.