Dealing with Server Crashes: The Hunger Games Survival Guide

In the online world your server is your lifeline and when it goes down and crashes so does your revenue, and productivity. The moment you notice your server is not responding panic sets in and Catholic school girls cringe at the words you use. Regardless of the reason your server went down, may it be an extreme traffic spike, or DDos Attack, the world feels like it is closing in on you. Don’t worry, I’m the Katniss of this online world and am here to get you out safely.

You and your dedicated server have fallen while running…

The first and easiest step to do when you have fallen is to pick yourself up. Assess the situation and if you can fix your server or solve the problem do it. Do not wait to see if the server will ‘pick itself up,’ instead figure out if you can fix it. If you can, you have just saved yourself from Tracker Jackers.

Scenario 1: You can’t pick yourself (and your server) up; you have broken something and are alone…

After you figure out that you can’t help yourself, don’t try and move. In other words, do not try and fix your system remotely especially if you don’t know how. If you do not have a managed server, and thus must take care of the problem yourself, wait until you are at least at the data center. You clearly must be onsite to repair any hardware problems. Trying to fix the problem remotely may cause more problems than anything else. As they say when you actual break something, do not try and move until someone can help you. Or in this case, wait until you can actually go to your data center.

If you cannot not make it to your data center right away nor can your sysadmin or IT personnel then contact your provider. You may be able to hire them to help fix your problem. Your provider may be able to provide you with a new machine or give you the best advice; after all this is what they do. Finding your own Rue to help in the meantime may be the difference between ‘life and death.’

Scenario 2: You can’t pick yourself (and server) up but you have allies…

When outsourcing data centers, some companies or people have found allies in their providers. If you have managed servers then contact your provider and tell them to fix it. Most likely your provider will have a failover solution to keep your website and data available. To keep with the metaphor, it is like your provider is camouflaging your server; you can survive hidden by the banks of the river for a while. Ask your provider to set up a new server or replacement VPS, shared hosting account or anything that will bring your site alive or at least announce that it is down due to maintenance. You never want a client to go on your site or try and access their information and find that it is down, contact them any way possible. May it be through Twitter, Facebook, or MockingJay, let your clients know you are experiencing technical difficulties and ask them to patiently stand by.

You are up and running or so it seems…

Once you have a new server or have repaired the old one, you still need to test it. Make sure that all the patches are completed, security firewalls are working, and that you have securely uploaded or recovered your data. It is like trying to run after being stung by Tracker Jackers, you are very likely to fall back down if you have not taken the appropriate precautions. If, after all your testing, everything appears to be working then you are ready to go online.

You have just survived an attack (or crash) and you are safe for now…

You decided to go back online and it seems that everything is back to normal. Make sure to write up an Incident Report or ‘official comment’ to send to all of your affected customers. Letting your customers know that you care about them by issuing an official sorry will help them trust you. Simple actions like ‘laying down flowers’ may be the difference from receiving loaf bread and losing your supporters. Your customers come first, show them that. After all, servers crash and it is how you handle it that makes all the difference.