Ryan C. Thompson's questions

I work on a CentOS 7 server where I frequently have to manually grant specific users access to specific files or directories. We do this using file ACLs, but I often run into an issue where I have set the ACLs on the file, but the user still cannot access it because they don't have permission to cd into the directory containing the file and/or any number of its parent directories. What follows is a tedious process of:

1. determining which specific directory or directories the user needs to be granted access to in order to actually give them access to the file; and
2. determining whether granting access to those directories might accidentally give them access to something else they aren't supposed to access, and making any appropriate adjustments to fix this.

Step 2 represents unavoidable complexity of the situation, but it should at least be possible to automate step 1. Specifically, given a user ID and a path to an existing file, is there a way I can get a full list of which directories along that path are not currently accessible to that user?

I can use crontab -e as a normal user to edit my user crontab. Does anacron have any similar feature for individual non-admin users, or does it only have a system-wide anacrontab?

Is there a way in linux to make "acl" part of the default mount options, so that it does not need to be specified individually for each filesystem in fstab?

When enabling userdirs (i.e. ~/public_html) in apache, what is the standard way to set up ownership and permissions of my public_html directory? Ideally I would like to give the apache server permission to do whatever it wants in that directory without making the whole thing world-writable. Is there a standard way to do this?

If the solution involves setting specific ownership or permissions, how can I ensure that any new files created in that directory will automatically receive the appropriate permissions?

I am running Ubuntu server 9.10, and I am trying to set up some commands in my crontab. However, it seems that none of these commands are executed. I have noticed that there is a crontab group. Do I need to be a member of this group in order for cron to like me?

I am using the following options in my ~/.ssh/config in order to enable connection multiplexing:

ControlMaster auto
ControlPath ~/.ssh/control/master-%r@%h:%p

However, this has the annoying problem that the first shell to connect to a particular server must be the last to disconnect, because it is the master connection that all the other connections are using. So if you log out of the master, it appears to just hang. To solve this, I would like to wrap ssh with a script that checks if the control master file exists, and if not, starts a master ssh process in the background. Then it would start a slave ssh session.

In order to accomplish this, my script would have to determine the path to the control file that ssh would use. This would entail parsing the ssh command line options and config files and implementing the logic for determining the ControlPath. Is there any way to just ask ssh what path it would use, so I can check it?