You must these server management tools. Without them, a crashed server likely means a stressful emergency migration.
While server crashes are always problematic they do not have to be business killers. There are three server tools you need to assure that a system administrator can help you recover your operations. If you don’t have these tools, you may want to check into your telekinetic powers because without them, you will often be facing a full server restore with potentially significant data loss.
Out-of-Band Management
What is It?
Quite simply it is an alternative means of getting to your server, network gear or other assets when the primary network has failed. In practice, you will find many different names for this solution: IPMI, DRAC, iLO, and KVM over IP.
Different server providers use different methods to deliver this. For example, ThePlanet sells Lantronic Spiders that provide KVM over IP access to your server. SoftLayer servers have integrated IPMI cards that deliver either a text console or KVM over IP. Since we deliver managed hosting, we use a centralize KVM over IP unit so we can easily manage all of our infrastructure assets.
How important is it?
Essential. This is the most important tool you can to your remote system administrator. Regardless of what the vendor labels the solution, full server management can be difficult without it. This tool basically puts me in front of the monitor. I can see the BIOS, watch the boot cycle and more. This is a critical tool. Without it, I would need to rely on my telekinetic powers which typically do not yield success.
Remote Power Control
What is It?
This allows me to pull the plug on your server remotely. With remote power control, you can reset or fully power the system on and off. In many cases, this can be sufficient to get a system going again.
How important is it?
Essential. Calling the data center or opening a ticket for a reboot can be a time consuming process, especially in larger facilities where simply walking to the server room may take 5 minutes.
When doing diagnostics, I may have to reboot a dozen times to fix something. If I am waiting 5-10 minutes for each reboot, an hour long outage can easily become a 2-3 hour outage. Also don’t forget you may be paying someone like me to sit around and wait while the data center reboots the server.
Rescue Mode
What is It?
Red Hat Linux has a “rescue mode’ that allows you to fix many common boot errors. If you had physical access to the server, you may use a rescue CD but for remote server management that may not be possible. To overcome this, many servers can boot over the network. Good dedicated server providers allow you to boot your server into a rescue mode simply by toggling something in their portal. This will put your server into a rescue state and allow a server administrator like myself to fix the issue.
How important is it?
Essential. I like Linux because you can often fix critical problems by using a rescue boot environment. I consider having access to a rescue environment key to building a robust server management framework. With a rescue system, I can often prevent full server restores.
Server Management Tool Summary
If you are leasing or buying a dedicated server, please make sure you give your system administrator the tools required to do the job. You would not ask your car mechanic to fix your car without a wrench, so don’t ask your sysadmin to go without these server management tools.
Spending a little more money to assure you have these 3 tools can reduce downtime and possible even save you from having to do a full system restore. I’ve had to order full restores for many simplistic issues simply because the server provider failed to deliver a robust remote server management environment. If you ever need help with your server, don’t forget we offer full Linux server management services.