I have a CentOS 6 machine that's hosting some non-critical Web applications.

I need it to be up in general, but a weekly late-night reboot won't cause any major business problems.

In order to keep the system up to date I put this in root's crontab...

6 3 * * * /usr/bin/yum update -y 2>&1 > /home/my_account/log/yum_update.log

So I'm running yum update -y daily there.

My main question -->

Sometimes an update requires a restart.

So to deal with that, I'd like to somehow set up this machine to reboot itself once per week.

What's a good way to do that?

(I considered adding a weekly cron job to run shutdown -r now, but I'm not sure if that would work well with a VPS machine.)

I understand many sysadmins would not consider this whole setup in general to be acceptable for a production system.

But I'm just hosting some personal apps and utils here. If an update goes awry and I need to restore from backup, it's ok.

My goal here is to keep this system patched and secure without costing too much of my time and attention which I need for other priorities that actually are critical.

In case it matters, I'll be running Apache, MySQL, and sshd.

Any help or advice would be greatly appreciated.

Thanks!

As root, add to your cron job

0 1 * * 1 /sbin/shutdown -r +5

Can't really see the advantage of this, if you're going to update daily (nightly) you probably want to reboot nightly too - otherwise a update that requires a reboot done on monday, will have to wait until sunday to get rebooted (plus stacking updates that require a reboot isn't a good idea either).

Of course willy nilly updates and reboots are just oh such a bad bad idea, but you already know that.

_________________
Either provide enough details for people to help, or sit back and listen to the crickets chirp.
Security thru obscurity is a myth - and really really annoying.

Thanks for the response, vonskippy.

Yeah, it isn't the greatest. Maybe I'll try to work out a better strategy, like manually once a week.

Realistically though, given the risk level and time constraints involved, this may be the least bad option for this particular host.

What does the "+5" option do?

Oh, I see. It gives logged in users (i.e., me) a five-minute warning. Cool

Will that work on linode, or will it result in a shutdown until Lassie detects the outage and schedules a boot from the management infrastructure?

_________________
Rgds
Stephen
(Linux user since kernel version 0.11)

It'll result in a shutdown until Lassie boots back up.

_________________
Paid support
How to ask for help
1. Give details of your problem
2. Post any errors
3. Post relevant logs.
4. Don't hide details i.e. your domain, it just makes things harder
5. Be polite or you'll be eaten by a grue

Yeah that's what the log is showing.

I thought maybe the usual shutdown command wouldn't work on Linode, which is why I originally asked.

It there a preferable way to configure the reboot?

Excepting kernel updates, no update on Linux should require a reboot; the files are overwritten in-place and then the service is restarted...

My bad for not trying it before posting.



So "why" doesn't this work on a Linode VPS when it works on a physical server?

_________________
Either provide enough details for people to help, or sit back and listen to the crickets chirp.
Security thru obscurity is a myth - and really really annoying.

Use the API? https://www.linode.com/api/linode/linode.reboot

_________________
Rgds

Stephen

(Linux user since kernel version 0.11)

I'm guessing it's something to do with how Xen is configured, but I dunno what. It does seem odd!

_________________
Rgds

Stephen

(Linux user since kernel version 0.11)

Key Lish with your ssh key, and then issue the reboot via Lish-via-ssh:

https://library.linode.com/using-lish-t ... public-key
https://library.linode.com/using-lish-t ... ish-tricks

ssh -t [manager-username]@lish-[location].linode.com [linode-name] reboot

-Chris

Excellent!

Thanks for the help, guys.