[Systems] Nagios container for SL

[Sam P.]

Hi Samuel,

On Fri, Aug 14, 2015 at 7:19 AM Samuel Cantero <scg at sugarlabs.org> wrote:

>
> ---------- Forwarded message ----------
> From: Samuel Cantero <scg at sugarlabs.org>
> Date: Sun, Aug 9, 2015 at 11:20 PM
> Subject: Nagios container for SL
> To: sysadmin at sugarlabs.org
>
>
> Hi all,
>
> I've set up a nagios container on freedom [1]. The user is nagiosadmin and
> the password is nagios. I will change the password if you decide to keep
> this service up. Please take a look.
>

It looks amazing.  Plus the notifications are useful.

We need to look at better auth methods though.  Using 1 username and
password is probably not the most secure method.  Maybe we could look into
a better password?  or client certs?

We could also do the auth at the nginx level on freedom.  If we wrap the
current cgi scripts in fast cgi so they can talk to nginx rather than
having apache proxied through nginx.  But that is not a biggie if we can
work the auth within apache.


>
> I have created three hostgroups: physical servers, docker containers and
> virtual machines. Advices are welcome.
>

I think that for the dockers that we need to check their cpu/memory usage.
That can be done by writing a plugin [1][2].  See
/srv/nagios-plugins/check_docker

[1]
https://gist.githubusercontent.com/ekristen/11254304/raw/adf549c5b663fe7127440f3c3b82d82cfc3c51c2/check_docker_container.sh
[2]
https://www.digitalocean.com/community/tutorials/how-to-create-nagios-plugins-with-bash-on-ubuntu-12-10


>
> I need information about what services we want to monitor on each vm
> inside justice and freedom. The main idea is not to replace munin. We just
> want to alert to the sysadmin team when a service goes down. Therefore
> sysadmin at sugarlabs.org will receive the nagios notifications unless you
> decide to change it to another email address.
>
> As you might guess, this won't work in case freedom goes down. If you like
> I can monitor if freedom is available through the nagios installed at my
> office and send an email to sysadmin at sugarlabs.org when freedom goes down.
>

You changed the freedom nagios to systems at lists because that got better
delivery, did you change your local one too?

Great idea setting up a monitor for all the machines though!  It was really
weird that we got no notification when freedom went down last time, so
hopefully this will stop that from happening again.


>
> I couldn't create a nagios subdomain for SL because I have no access to
> ns1.slo. Anyway I have added the nagios configuration for nginx.
>


Sorry.  My raspberry pi has given up on me, so I can't access a (way
outdated) version of dnssec-signzone that will work with the nameservers.
Anybody else able add nagios CNAME freedom?

Thanks,
Sam


>
> Samuel
>
> [1] http://18.85.44.59:8081/nagios3/
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sugarlabs.org/private/systems/attachments/20150814/c285afb3/attachment.html>