nadnerB's CloudatCost Project Journal

nadnerB

@nadnerB said:

Hmmm, perhaps this is not a good idea...
Comments @JaredBusch, @thecreativeone91, @scottalanmiller or @thanksajdotcom ?
EDIT: I'll hold off on deploying this for now.

I log in as root directly to all my servers.

Thanks for your input but I won't be doing this

nadnerB

@thecreativeone91 said:

@nadnerB said:

Hmmm, perhaps this is not a good idea...
Comments @JaredBusch, @thecreativeone91, @scottalanmiller or @thanksajdotcom ?
EDIT: I'll hold off on deploying this for now.

You don't edit the file You'd gpasswd -a nadnerb wheel where nadnerb is the username you wish to give sudo privileges too.

Fantastic! Thanks!

nadnerB

@thecreativeone91 said:

@thanksajdotcom said:

@nadnerB said:

Hmmm, perhaps this is not a good idea...
Comments @JaredBusch, @thecreativeone91, @scottalanmiller or @thanksajdotcom ?
EDIT: I'll hold off on deploying this for now.

I log in as root directly to all my servers.

I personally would disable root access over SSH after the initial setup.

On the secret To-Do list

@nadnerB said:

@thecreativeone91 said:

@nadnerB said:

Hmmm, perhaps this is not a good idea...
Comments @JaredBusch, @thecreativeone91, @scottalanmiller or @thanksajdotcom ?
EDIT: I'll hold off on deploying this for now.

You don't edit the file You'd gpasswd -a nadnerb wheel where nadnerb is the username you wish to give sudo privileges too.

Fantastic! Thanks!

No Problem. It's just a group you add it to, as the group has sudo premissions (sudoers file) .

scottalanmiller

@thanksajdotcom said:

I log in as root directly to all my servers.

Why?

scottalanmiller

@nadnerB said:

Righto, so it looks like the Sudoers file, that I need to edit, is read only.

Just means you have to tell the editor that you "mean it" when you save. In vi that means :w! instead of :w

nadnerB

Righto, I've blocked root access via SSH and renamed the server to something more useful (for ron... later on)

nadnerB

Hmmm, attempting to install htop is proving to be more difficult than yum -y install htop.
I can't seem to connect to any of the mirrors.
*http://mirror.netflash.net/centos/7.0.1406/updates/x86_64/repodata/repomd.xml: [Errno 14] curl#6 - "Could not *resolve host: mirror.netflash.net; Unknown error"
Trying other mirror.

scottalanmiller

Often that means that DNS isn't set up. Can you lookup addresses in general?

You might need to set dns in resolv.conf

coliver

If the DNS issue doesn't resolve it you may have to refresh your YUM cache. I think a yum -clean all or yum -clean headers will do that.

nadnerB

Excellent suggestions! I'll check that out when I get home

nadnerB

Found the location of resolv.conf and how to edit here: http://ask.xmodulo.com/configure-static-dns-centos-fedora.html
htop now installed. I quite like it

scottalanmiller

All configuration is just in /etc

nadnerB

@scottalanmiller nice. I screen shot the directory from my laptop with a GUI

EDIT: that sounded disturbingly like TV IT. I apologise to all those who are racing for the spew buckets

nadnerB

Reading for when I get home: https://www.centos.org/docs/5/html/Deployment_Guide-en-US/ch-vpn.html

scottalanmiller

That's from CentOS 5. A bit old these days.

I thought just about everyone had given up on C@C now. Espcially since they never said what they'd do to fix the problems or had their webinar thing they promised.

coliver

@thecreativeone91 said:

I thought just about everyone had given up on C@C now. Espcially since they never said what they'd do to fix the problems or had their webinar thing they promised.

Nope still using them for personal lab stuff... still really can't beat the price for things that aren't really important.

scottalanmiller

@thecreativeone91 said:

I thought just about everyone had given up on C@C now. Espcially since they never said what they'd do to fix the problems or had their webinar thing they promised.

Still using and increasing usage here. For lab boxes they remain unbeatable. For production, we never were looking at them for that. I was pretty shocked when people started trying to deploy production to them. A little testing showed that normal production workloads weren't going to work there.