Email UsCloudManage: Cloud Hosting Automation at best
Posted: March 30, 2011 at 7:10 pm | No CommentsIf you’re in the technology industry and especially if you run any critical services, the last thing you want to do is contact support. Let’s face it, with the best possible customer service, the last thing you want to be on is on hold while your boss reminds you that the server is down (or rather, the internet is down). While developing the next generation of CloudManage control panel that has already been released, this was an extremely important focus of ours.
We all love the ATMs, could you imagine waiting in line every time you wanted a quick $20 to use?! The great thing about them is they allow you to perform a task instantly and automatically without the presence (or having to deal with a human) — Sounds so futuristic, but it makes everything so much simpler! As banks started to realize how useful people find ATMs (and also, they reduce lines at banks) — They started to include more features, so the typical ATM went from one button of Withdraw to Deposit, view balance, etc.
After countless hours of meetings, we got a pretty decent list of tasks that server administrators would love to perform on their own, without the requirement of contacting their vendor or support, so that they can get back up as soon as possible.
It’s 3:37 AM, your boss calls you to remind you that the 12 free website monitoring tools that he setup all report that the internet is down, and someone has to fix it! Half-asleep, you stumble to your laptop and try a few pings, connections and right as expected, the server is not working!
As the cloud server is a virtual server, you’d need to contact your cloud hosting customer support, at 3 AM, to ask them what is going on. However, on our platform, we do offer a remote console, think of it as your personal teleporter at 3AM instantly right in front of your machine. Regardless if your server has internet access or not, the console will always work, it is the equivalent of connecting your monitor to the server directly.
So, the kernel crashed huh? You write down the errors that were raised, now you can’t even login to the server to restart it. No worries, with a fancy swipe of a selector, your server is turned off. Also, our system is smart enough to detect the current state of the server and attempt a soft shutdown, and if that doesn’t work out — well, we bring the heavy weaponry and forcingly shut off the server.
Another swipe, the server is restarted and running again. However, it’s still not pingable, the sun is starting to shine and you could really use some sleep right now. At this point, if it wasn’t for our magical console, you’d need to contact the support team for them to see what’s up, however, one click at “Server Console” and you can see that the server is simply running a file system check for consistency, so you can follow it until it’s complete, make sure everything is up, tell your boss that the “Internet” is working again, and hit the sack.
Now, all of this could probably take from 1 to 3-4 hours if you had to keep going back and forth with support, the whole duration of this without contacting support takes you a maximum of 15 minutes and allows you to take full control of the server.
This is just one of the great automation systems that we put in place, there are tons more that we implemented in order to reduce support contact time and increase automation.
Protip: Not only does the server console help you access the server when it’s crashed, but changing the servers boot parameters to include “single” will boot the server in single user mode to run any file system checks, bet your provider doesn’t allow you that. 
