CentOS 6.5 constantly losing commands

Bash: Command not found

I read up in one situation and navigated directly to the service I wanted to start and did it. Another time I had to type >PATH=$PATH:./sbin and that worked.

Now I can't start services again and I am just getting sick of it.

IS there a permanent fix for this?


typical losses are "service" but other things have not been found as well.

What is PATH set to when you log in? It sounds like it isn't set.

when i checked it searched usr/bin and lib folders.

I think i found my problem.

I exited out of terminal and did not su - on reentry. sinc ethe server has not been restarted each terminal opening was the same one, so got comfortable and then when i actually typed exit, it logged out. Not used to that


doh!!


Next step is to figure out why I can't remote in as a normal user anymore.

to diagnose these issues you may want to do something like setup a specifc ssh/sftp log instead of having it dump into /var/log/messages or /var/log/secure

To do this, add something like this to your rsyslog.conf


(restart rsyslog after adding this)

This will at least isolate the specific messages you are looking for. In addition, to trouble shoot you should have windows open tailing secure, messages and the ssh log so that as you make attempts you can see what messages are happening real time.

(tail -f <file name>) for real time tailing

Will logging everything increase overhead much? It is a core2 duo 64xx and only 4GB of RAM.

Going to do a reinstall this weekend and will set up logging AND going to write down everything I do as well.

Came in to a kernel panic this morning after remote restart last night. Going to run some hardware stressing first just to make sure. then reinstall as a server base, since I know how to do enough CLI now to get X running.

although all the guides I am reading contradict each other on how to setup stuff


Server with Samba, NFS shares, rsync, LAMP, phpmyadmin and XRDP.

Neuromancer said:

Will logging everything increase overhead much? It is a core2 duo 64xx and only 4GB of RAM.

Click to expand...

The information is already being logged. My suggestion is to simply split it into different files to reduce troubleshooting complexity (potentially). All the rsyslog line above does is say "If you encounter sshd as a syslog tag, log it to the location specified instead of the generic location"

Great, thanks again!

I assume you mentioned ssh because it is the root command for XRDP? (I think I recall seeing it in the console login script)

Neuromancer said:

Great, thanks again!

I assume you mentioned ssh because it is the root command for XRDP? (I think I recall seeing it in the console login script)

Click to expand...

No I didnt know/am unfamiliar with XRDP. Usually when people say remote access is not working for a linux box, they are referring to ssh