offsides

This appears to be caused by the STORABLE_freeze functions, and the fact that the 3 sub-packages (Readonly::Scalar, Readonly::Hash and Readonly::Array) are contained within Readonly.pm and not separate packages (the error...

Further investigation and testing makes me wonder what the point of the STORABLE_freeze functions is, and if they were ever tested. Because commenting them out doesn't seem to break anything,...

Here's the service definition as well as the log output: ``` apply Service "Linux Update Status" { check_command = "systemd_service" assign where host.vars.os == "Linux" && host.vars.agent_endpoint command_endpoint = host.vars.agent_endpoint...

Just another tidbit - I ran strace against my running icinga2 processes following forks, and don't find the string 'test_event' anywhere, making me think that it's never actually being sent....

It looks like this may be related to #9259, which looks like it's designed to make sure commands don't run on both masters in a HA situation. But it may...

I will have to test that, but right now I'm working on putting a workaround in place for handling the triggers since this didn't appear until I was already in...

Interesting - disabling mod_vroot gives me the full path for %f, but not for %w (e.g., %w shows /old_filename, whereas %f shows /path/to/new_filename). I'm not sure what other effects disabling...

This is for both ftp and sftp sessions. The results are identical for with and without mod_vroot - without mod_vroot the new filename shows the full path, while the old...

Not sure if I did something wrong, but it doesn't appear to be working for me. I'm getting relative paths for both old and new filenames with mod_vroot enabled. I...

mod_vroot 0.9.12. Relative means I get the path as it would be seen inside the chroot - i.e., if the absolute path is '/data/ftp/ftp_test/filename', and the chroot is into /data/ftp/ftp_test,...