A few configuring questions with regard to services like vsftpd

davidstvz · 08-20-2008, 12:57 PM

I assume that after I change the .conf file for vsftpd, I need to stop and start the service again. How is that done?

By the way, instead of learning the proper way, I tried something stupid that seems to have broke things a little bit.

I tried to kill the vsftpd process and then run "/usr/sbin/vsftpd &" again (as initially instructed by a certain website (vsftpdrocks.org) hoping that would restart it. Instead it tells me:

and I got:

[2] 1394
[1] Terminated /usr/sbin/vsftpd

Anyway, someone please give me an idea of how to do this properly, and let me know if my little process killing thing will require any cleaning up or a reboot (I have no idea what the above output means).

davidstvz · 08-20-2008, 12:58 PM

As a quick follow up, although it didn't look pretty (as per the above output) it apparently worked. User/password ftp is now working on the machine for me.

Still, I figure there is a better way than issuing a kill command on the vsftpd process.

trickykid · 08-20-2008, 01:00 PM

If it's configured for standalone mode which should be indicated in your conf file, that should be the proper way usually to start vsftpd. If not in standalone mode, usually it's invoked through inetd or maybe another init script, depending on what distribution you are using.

trickykid · 08-20-2008, 01:02 PM

Quote:

Originally Posted by davidstvz

Still, I figure there is a better way than issuing a kill command on the vsftpd process.

If you don't have a script to start and stop the services, write one and simply make the killing of it do it in a nice way, like kill -15 (SIGTERM) which will allow the service/daemon to cleanup any leftovers before completing dying. Kill -9 should be avoided in most cases, it's the last case effort if all else fails when you're trying to kill a process.

davidstvz · 08-20-2008, 01:58 PM

Quote:

Originally Posted by trickykid

If you don't have a script to start and stop the services, write one and simply make the killing of it do it in a nice way, like kill -15 (SIGTERM) which will allow the service/daemon to cleanup any leftovers before completing dying. Kill -9 should be avoided in most cases, it's the last case effort if all else fails when you're trying to kill a process.

Ah, so killing isn't a bad idea, I just have to do it the right way.

I am currently running standalone mode, but I'm considering the xinetd method.

One more question, what does the & symbol mean as an option after starting vsftpd?

Example:

/usr/sbin/vsftpd &

?

davidstvz · 08-20-2008, 02:58 PM

EDIT:

The problem below been explained to me now in another thread.

-------------------------------------------

OK, I restarted xinetd but it "failed"

My /var/log/messages says this:

Code:

Aug 20 14:36:57 serv2 xinetd[7180]: bind failed (Address already in use (errno = 98)). service = ftp
Aug 20 14:36:57 serv2 xinetd[7180]: Service ftp failed to start and is deactivated.
Aug 20 14:36:57 serv2 xinetd[7180]: 7180 {init_services} no services. Exiting...

davidstvz · 08-20-2008, 03:10 PM

EDIT: This has been addressed too. Thank you (my post, two above this about the & symbol still stands unanswered though!)

----------------------------

I have reverted the config files to their earlier form, but according to the logs (as best as I can tell) xinetd is still trying to start vsftpd

(I have disabled set to yes, and in the vsftpd.conf file the thing is set to standalone)

I backed up and deleted the vsftpd file in /etc/xinetd.d/ and it still won't start.

This is really starting to worry me! I need this server up and running properly within a few days and I'm not sure what the problem is here.