Do you have a web site disaster plan in order?
I am betting you likely don’t.
Why is a disaster plan important?
The unknown is ever present in the world of technology. With the rise of malware and CPU defects, the chances of your site going down by unseen forces is getting higher every day. You literally could wake up one morning and your site is no longer online, or worse it is being held for ransom. Add into the mix the number of web hosting companies that go out of business or are sold to another company. If you don’t have a worse case disaster plan in place, it is my opinion you are not doing yourself any favors. It is very easy to put together and can be accomplished by anyone. This would be like having an emergency go bag if you live in a earthquake zone.
What are some key things you need to have on your disaster plan?
Login details for your Domain and where it is registered (username, password, phone number and support email address).
It may or may not be registered with the same company that hosts your website. I would make a document that includes your login details, contact phone number and support email address. Put this along with the others we will be covering into a envelope and seal it, then put that in a safe place.
Login details for your hosting account (username, password, phone number and support email address).
This is the location where your website is actually being served from. Put this information in the same envelope as the rest of the ones we are covering. It is also important to have a phone number and support email address along with your login details.
A current backup or archive.
We have discussed this several times on this podcast. You should have a current backup or archive you can work with of at least your website, and possibly of your whole hosting account. If you have been backing up externally or manually copying to a local disk drive, put this information and location of the backup in the envelope with the other information.
Now that you have your login details sorted out, you need to have some basic DNS information. I personally like to have a complete zone listing of all of my DNS entries. These are things like;
- What are my nameservers and where are they pointing? Nameservers are vital to knowing where your zone record is being kept. If your nameservers vanish, your domain vanishes from the internet.
- Where does www and yourdomain.com point to?
- What are my MX records?
- Do I have a custom record that is used for connecting to my mail server? For example, do you use mail.yourdomain.com and if so where is it pointing too?
- Are there any other records I need for my site to be online? Custom records for a cdn, custom txt records that have been added, SPF records? There are many types of records that can be added to DNS. Some of them are for email, some are for proving you own a domain (google validation comes to mind). All records should be tracked and kept with your disaster plan records. You never know when you may need to recreate a zone entry.
Success plan not unlike the disaster plan.
What happens if your site starts getting a large amount of traffic. Good for you, bad for your hosting company if your on shared hosting. I have seen this type of thing happen time and time again. A article you may have written, or a product you are offering gets picked up by national news or celebrity likes your product. This is great news for you, but this can often result in your site going down or even being taken offline by your hosting company. How do you deal with a “scuccess” hit often involves the same things as a disaster plan. You may find yourself needing to move to a new host rather rapidly. Have those contact information and login details at the ready in your disaster plan packet. Lets just call this the “What if” packet.
If you are just experiencing some temporary increased traffic, meaning you don’t think it will last for very long as the hype dies down. There are a few steps you can do to help with the site traffic increase, which will likely help with server load.
- Use a caching service like cloudflare. We have discussed this in the past. Basic cloudflare services are free and it only takes a minute to setup. This will act as a buffer between your host and the people trying to access your site.
- Make sure you use expires and headers so files are cached. Another topic we have discussed in the first episode.
- Make sure you are compressing the site files with mod_deflate. See episode 1 for more details. Or listen to the end of this episode for the quick tip.
- Enable a caching plugin in your framework. Something like wp super cache or w3 total cache for wordpress will save you a lot of headaches with a sudden spike in site traffic. This will also lower server load by reducing the mysql queries required to load your site by making some of the site pages almost static in nature. This will in turn keep your host happy. This is not the same as cloudflare caching service.
- Serve a static site during the increase in traffic. This one is a little more tricky, but it is definitely possible. By removing the need to have mysql and php render pages, your site will load faster and have almost zero load on the server. This requires planning ahead however and having static pages ready to go.
- Work with your hosting provider to see if you can to keep your site online. If they are less than helpful, then reach out to the world and get a recommendation for a new host. A good host will want you to grow and be a part of your growth process. If they just suspend your account because you are successful suddenly, then they are impeding your growth and should be removed from the equation. If the host offers some suggestions to you, no matter if they sound complicated, and want to work with you in providing even a temporary solution to the situation, then you should listen and see if they can help.
Things to NOT do. Do not allow your host to move you to a tiny VPS of your own. This is the number one thing I see and it will kill your site, but save your hosts butt. If your site is already creating a problem on a very large shared servers with possibly many CPU cores and many Gigs of ram, what good is moving you to a 1 core and 1 gig of ram VPS going to do. They just want you off their shared server as fast as they can, they are not offering a solution but passing the buck to you and making a few bucks in the process. You site will never stay online in a small VPS unless you have someone that you can call on to make massive tweaks to the VPS itself, install specific software and configure it, this often requires a system administrator/engineer to do.
Do NOT try and block the inbound traffic that is being generated, this includes changing the URL, blocking IPs in .htaccess or server firewall. You want that traffic to come in, if there are elements on that page that require external resources, like a facebook or twitter feed, remove that code during the spike in traffic. These can potentially slow down your page speed.
The biggest take away I want to share with everyone is to be proactive and not reactive. Whether it is a disaster plan or a success plan, the “what if” scenario should be on the minds of everyone. And if you are not ready for it, it can be devastating to your site, your finances and even your emotional state. Like any other disaster preparedness scenario, regaining control of the situation as fast as possible will allow you to continue on with your life. It will remove stress and worry. If you get an email from your hosting provider saying, “your site has been shutdown because….” you will know how to proceed because of your planning. Take some time out of your busy week and determine the best way to handle your “what if” scenario, it will make your life a lot better. If you have already put together a “what if” packet, then please share your experience and tips you may have with me. I would love to hear about them.
Quick tip today is gzip compression in cPanel, you can also see a video I did on this here.