Alejandro Colomar

If the sysadmin wants to limit the ability of unitd to create certain directories, it can limit the permissions of the user that will run it, so that it's not...

> I found some information I didn't know before. #742 > > ``` > alx@asus5775:~/src/nginx/unit$ sudo /opt/local/unit/sbin/unitd > 2022/08/05 20:12:22 [alert] 21613#21613 bind(6, unix:/opt/local/unit/var/run/unit/control.unit.sock.tmp) failed (2: No such file or...

> Let's separate it into two requirements. > > 1. If it reasonable to limit the creation of socket directory? For example: > > > ``` > { > "listeners":...

> > We're not (yet) modifying that, right? @ac000 this change doesn't affect listeners, right? > > For devs, `nxt_socket_bind()` is a generic function, we can't predict who uses it...

> Just to wonder if these automatically created directories need to be cleaned up after unitd quits? No, they need not be cleaned up. The OS nukes `/run` at each...

I'm having some doubts:

Regarding the trailing directories: it's a minor issue, especially in the standard path, which is `/run`, since it's nuked at reboot. One might say that in non-standard paths that may...

I applied a fix for a bug, that I think could fix your bug. Could you test building unit from source and check if you still have the bug? Thanks!

Hi! I'm trying to solve this in a much simpler way: . I added a commit that I expected should unlink the sockets after use, but I'm getting EACCESS, and...

Hi Evgenii! > > Hi! > > I'm trying to solve this in a much simpler way: #655. I added a commit that I expected should unlink the sockets after...