Moz Q&A is closed.
After more than 13 years, and tens of thousands of questions, Moz Q&A closed on 12th December 2024. Whilst we’re not completely removing the content - many posts will still be possible to view - we have locked both new posts and new replies. More details here.
How can I prevent duplicate pages being indexed because of load balancer (hosting)?
-
The site that I am optimising has a problem with duplicate pages being indexed as a result of the load balancer (which is required and set up by the hosting company).
The load balancer passes the site through to 2 different URLs:
Some how, Google have indexed 2 of the same URLs (which I was obviously hoping they wouldn't) - the first on www and the second on www2.
The hosting is a mirror image of each other (www and www2), meaning I can't upload a robots.txt to the root of www2.domain.com disallowing all. Also, I can't add a canonical script into the website header of www2.domain.com pointing the individual URLs through to www.domain.com etc.
Any suggestions as to how I can resolve this issue would be greatly appreciated!
-
There are two ways to handle load balancing, and it appears that your hosting company / server company chose to use the DNS round-robin routing option.
According to the Wikipedia page on load balancing:
http://en.wikipedia.org/wiki/Load_balancing_(computing)"Load balancing usually involves dedicated software or hardware, such as a multilayer switch or a Domain Name System server process."
Round Robin DNS Load Balancing: Basically you use the DNS routing system to handle requests. When someone visits your site, 50% of the people are routed to www.domain.com, and 50% are routed to ww1.domain.com. Both sites contain the same identical content; it's the URLs that are slightly different. Sometimes the domains are the same; but you have different IP addresses for www.domain.com.
Advantages: you don't need a dedicated load balancing piece of software or hardware, so it's less expensive.
Disadvantages: this technique exposes the individual web servers to the end user seeing the site. You can also suffer from duplicate content penalties, too. Finally, if you are relying on the round robin DNS system for load balancing, and a DNS server or one of the Web servers goes down, there's not an easy fail-over (as many DNS records are cached).More about Round Robin DNS: http://en.wikipedia.org/wiki/Round-robin_DNS
Hardware / Software Load Balancer:
In this case, your DNS zone file tells the end user to go to one IP address when they type in www.domain.com. The hardware or software load balancer then sees the request, and then hands off the content to one of the web servers in a cluster.Advantages: No duplicate content penalty; to the end user, they just see one web server and not individual sub-domains (www.domain.com and ww1.domain.com). A load balancer can also cache specific items like a CSS page, so the load on the Web server is even more minimal.
Disadvantages: You're introducing another piece of hardware or software (i.e. more cost); this piece could also be a single point of failure into the mix. You need someone to figure out how to set this up and make sure it all works.
More on this type of Load Balancing: http://en.wikipedia.org/wiki/Load_balancing_(computing)#Internet-based_services
Load balancing can get complicated as soon as you have databases involved, but with a good design, multiple front end Web servers can talk to one single backend database server. The goal would be to cache as much content as possible as "static" elements, using caching systems like Varnish, that essentially turn database-driven pages into static, old-school HTML pages. And then only when someone needs to save something from the database (i.e. making a purchase on an eCommerce site), the system then interacts with it.
My recommendation:
(1) Move from the Round Robin Robin DNS to a hardware or software load balancer.(2) If that isn't an easy solution, implement the Round Robin DNS solution to use identical A records for each server.
For example, you might have identical entries in your DNS zone files for both DNS servers:
www.domain.com A 69.94.15.10
NS2.domain.com:
www.domain.com A 75.64.18.12This should at least eliminate your duplicate content issue, but you still do have a few disadvantages (described above). This also could lead to server issues, as the servers might be confused if they are the authoritative ones.
And if both servers are sending email, pay special attention to your SPF record, to make sure that you are allowing both IP addresses to be able to send email. (This is often overlooked.)
Hope this is helpful!
-- Jeff
Got a burning SEO question?
Subscribe to Moz Pro to gain full access to Q&A, answer questions, and ask your own.
Browse Questions
Explore more categories
-
Moz Tools
Chat with the community about the Moz tools.
-
SEO Tactics
Discuss the SEO process with fellow marketers
-
Community
Discuss industry events, jobs, and news!
-
Digital Marketing
Chat about tactics outside of SEO
-
Research & Trends
Dive into research and trends in the search industry.
-
Support
Connect on product support and feature requests.
Related Questions
-
Password Protected Page(s) Indexed
Hi, I am wondering if my website can get a penalty if some password protected pages are showing up when I search on google: site:www.example.com/sub-group/pass-word-protected-page That shows that my password protected page was indexed either before or after adding the password protection. I've seen people suggest no indexing the page. Is that the best method to take care of this? What if we are planning on pushing the page live later on? All of these pages have no title tag, meta description, image alt text, etc. Should I add them for each page? I am wondering what is the best step, especially if we are planning on pushing the page(s) live. Thanks for any help!
Intermediate & Advanced SEO | | aua0 -
Home page suddenly dropped from index!!
A client's home page, which has always done very well, has just dropped out of Google's index overnight!
Intermediate & Advanced SEO | | Caro-O
Webmaster tools does not show any problem. The page doesn't even show up if we Google the company name. The Robot.txt contains: Default Flywheel robots file User-agent: * Disallow: /calendar/action:posterboard/
Disallow: /events/action~posterboard/ The only unusual thing I'm aware of is some A/B testing of the page done with 'Optimizely' - it redirects visitors to a test page, but it's not a 'real' redirect in that redirect checker tools still see the page as a 200. Also, other pages that are being tested this way are not having the same problem. Other recent activity over the last few weeks/months includes linking to the page from some of our blog posts using the page topic as anchor text. Any thoughts would be appreciated.
Caro0 -
How can I get Bing to index my subdomain correctly?
Hi guys, My website exists on a subdomain (i.e. https://website.subdomain.com) and is being indexed correctly on all search engines except Bing and Duck Duck Go, which list 'https://www.website.subdomain.com'. Unfortunately my subdomain isn't configured for www (the domain is out of my control), so searchers are seeing a server error when clicking on my homepage in the SERPs. I have verified the site successfully in Bing Webmaster Tools, but it still shows up incorrectly. Does anyone have any advice on how I could fix this issue? Thank you!
Intermediate & Advanced SEO | | cos20300 -
Do internal links from non-indexed pages matter?
Hi everybody! Here's my question. After a site migration, a client has seen a big drop in rankings. We're trying to narrow down the issue. It seems that they have lost around 15,000 links following the switch, but these came from pages that were blocked in the robots.txt file. I was wondering if there was any research that has been done on the impact of internal links from no-indexed pages. Would be great to hear your thoughts! Sam
Intermediate & Advanced SEO | | Blink-SEO0 -
Do I need to re-index the page after editing URL?
Hi, I had to edit some of the URLs. But, google is still showing my old URL in search results for certain keywords, which ofc get 404. By crawling with ScremingFrog it gets me 301 'page not found' and still giving old URLs. Why is that? And do I need to re-index pages with new URLs? Is 'fetch as Google' enough to do that or any other advice? Thanks a lot, hope the topic will help to someone else too. Dusan
Intermediate & Advanced SEO | | Chemometec0 -
Date of page first indexed or age of a page?
Hi does anyone know any ways, tools to find when a page was first indexed/cached by Google? I remember a while back, around 2009 i had a firefox plugin which could check this, and gave you a exact date. Maybe this has changed since. I don't remember the plugin. Or any recommendations on finding the age of a page (not domain) for a website? This is for competitor research not my own website. Cheers, Paul
Intermediate & Advanced SEO | | MBASydney0 -
No-index pages with duplicate content?
Hello, I have an e-commerce website selling about 20 000 different products. For the most used of those products, I created unique high quality content. The content has been written by a professional player that describes how and why those are useful which is of huge interest to buyers. It would cost too much to write that high quality content for 20 000 different products, but we still have to sell them. Therefore, our idea was to no-index the products that only have the same copy-paste descriptions all other websites have. Do you think it's better to do that or to just let everything indexed normally since we might get search traffic from those pages? Thanks a lot for your help!
Intermediate & Advanced SEO | | EndeR-0 -
De-indexing product "quick view" pages
Hi there, The e-commerce website I am working on seems to index all of the "quick view" pages (which normally occur as iframes on the category page) as their own unique pages, creating thousands of duplicate pages / overly-dynamic URLs. Each indexed "quick view" page has the following URL structure: www.mydomain.com/catalog/includes/inc_productquickview.jsp?prodId=89514&catgId=cat140142&KeepThis=true&TB_iframe=true&height=475&width=700 where the only thing that changes is the product ID and category number. Would using "disallow" in Robots.txt be the best way to de-indexing all of these URLs? If so, could someone help me identify how to best structure this disallow statement? Would it be: Disallow: /catalog/includes/inc_productquickview.jsp?prodID=* Thanks for your help.
Intermediate & Advanced SEO | | FPD_NYC0