Why are there lots of 404s after setting up CDN?
-
I just setup Cloudfront CDN through W3 Total Cache. Everything looks good but there is one problem that I have encountered:
After activating the CDN none of the images are available at the older image URLs and they are throwing a 404 error.
Let me give you an example for this:
1. Before I setup the CDN, let's say an image was available at http://example.com/wp-content/uploads/2015/03/leap-of-faith.jpg
2. After I setup the CDN, the image is available at http://cdn.example.com/wp-content/uploads/2015/03/leap-of-faith.jpg and the good part is the URLs in the blog posts where this image was attached is updated to reflect the above mentioned URL. But the problem is that when visit the older URL of the image (which is what Google has crawled earlier, I get a 404 error).
Can you help me how to avoid this problem?
Ravi C
-
Thanks Dirk.. That sounds good.
-
Hi,
Do you get a lot of traffic coming through image search? Most of the images you use on the site seem to be stock images, so normally the % of image search traffic shouldn't be that big.
If you receive limited or no search traffic from image search, you don't really have to do anything special. There 404 errors in WMT will disappear after a while & the new images will get indexed. Normally the 404's will have no impact on search traffic.
If all CDN's contain all the images, you could always redirect the original image folder to one of the cdn's - but it not strictly necessary.
rgds
Dirk
-
I would say you may look at the set up process of CDN may not be as per the required criteria. IF you just check it thoroughly you may be able to get rid of it.
-
Thanks for that reply Dirk.
I think what you are referring to is quite applicable when the CDN is setup via using a S3 bucket. I followed the following guide to setup my CloudFront CDN:
https://www.doitwithwp.com/set-up-w3-total-cache-with-amazon-cloudfront-cdn
Here are the 2 problems that I'm facing currently:
1. The images appear at multiple CDN locations - http://cdn5.sarkarilife.com/wp-content/uploads/2015/01/bank-awareness-gk-ibps-bank-exams.jpg as well as http://cdn1.sarkarilife.com/wp-content/uploads/2015/01/bank-awareness-gk-ibps-bank-exams.jpg .
2. The same image is not available at the original location - http://sarkarilife.com/wp-content/uploads/2015/01/bank-awareness-gk-ibps-bank-exams.jpg
Looking forward to your response.
-
Hi,
You could put a 301 redirect of the old image folder to the new location. The easiest alternative is to keep the images in both places for a while, until Google has indexed the new location (which can take a few weeks/months). Normally, if all the internal links have been updated, there should be no links to old location, so these images will disappear from the index and replaced by the ones in the new location. Once they are indexed on their new location, you can delete them in the old location
rgds,
Dirk
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
-
Is there a best practice for how to set up an Age Restriction prompt for a website (vape equiptment)
I am working on a website that sells vape equipment and supplies and is required to have an age restriction (birthday) prompt in order to enter the site. Is there a 'best practice' way to implement this?
Web Design | | RandolphMorris0 -
Lots of Listing Pages with Thin Content on Real Estate Web Site-Best to Set them to No-Index?
Greetings Moz Community: As a commercial real estate broker in Manhattan I run a web site with over 600 pages. Basically the pages are organized in the following categories: 1. Neighborhoods (Example:http://www.nyc-officespace-leader.com/neighborhoods/midtown-manhattan) 25 PAGES Low bounce rate 2. Types of Space (Example:http://www.nyc-officespace-leader.com/commercial-space/loft-space)
Web Design | | Kingalan1
15 PAGES Low bounce rate. 3. Blog (Example:http://www.nyc-officespace-leader.com/blog/how-long-does-leasing-process-take
30 PAGES Medium/high bounce rate 4. Services (Example:http://www.nyc-officespace-leader.com/brokerage-services/relocate-to-new-office-space) High bounce rate
3 PAGES 5. About Us (Example:http://www.nyc-officespace-leader.com/about-us/what-we-do
4 PAGES High bounce rate 6. Listings (Example:http://www.nyc-officespace-leader.com/listings/305-fifth-avenue-office-suite-1340sf)
300 PAGES High bounce rate (65%), thin content 7. Buildings (Example:http://www.nyc-officespace-leader.com/928-broadway
300 PAGES Very high bounce rate (exceeding 75%) Most of the listing pages do not have more than 100 words. My SEO firm is advising me to set them "No-Index, Follow". They believe the thin content could be hurting me. Is this an acceptable strategy? I am concerned that when Google detects 300 pages set to "No-Follow" they could interpret this as the site seeking to hide something and penalize us. Also, the building pages have a low click thru rate. Would it make sense to set them to "No-Follow" as well? Basically, would it increase authority in Google's eyes if we set pages that have thin content and/or low click thru rates to "No-Follow"? Any harm in doing this for about half the pages on the site? I might add that while I don't suffer from any manual penalty volume has gone down substantially in the last month. We upgraded the site in early June and somehow 175 pages were submitted to Google that should not have been indexed. A removal request has been made for those pages. Prior to that we were hit by Panda in April 2012 with search volume dropping from about 7,000 per month to 3,000 per month. Volume had increased back to 4,500 by April this year only to start tanking again. It was down to 3,600 in June. About 30 toxic links were removed in late April and a disavow file was submitted with Google in late April for removal of links from 80 toxic domains. Thanks in advance for your responses!! Alan0 -
Could our drop in organic rankings have been caused by improper mobile site set-up?
Site: 12 year old financial service 'information' site with lead gen business model. Historically has held top 10 positions for top keywords and phrases. Background: The organic traffic from Google has fallen to 50% of what it was over the past 4 months compared to the same months last year. While several potential factors could be responsible/contributing (not limited to my pro-active removal of a dozen old emat links that may be perceived as unnatural despite no warning), this drop coincides with the same period the 'mobile site' was launched. Because I admittedly know the least about this potential cause, I am turning to the forum for assistance. Because the site is ~200 pages and contains many 'custom' pages with financial tables, forms, data pulled from 3rd parties, custom/different layouts we opted for creating a mobile site of only the top 12 most popular pages/topics just to have a mobile presence (instead of re-coding the entire site to make it responsive utilizing a mobile css). -These mobile pages were set up in an "m." subdomain. -We used bi-directional tagging placing a rel=canonical tag on the mobile page, and a rel=alternate tag on the desktop page. This created a loop between the pages, as advised by Google. -Some mobile pages used content from a sub page, not the primary desktop page for a particular topic. This may have broken the bi-directional 'loop', meaning the rel=canonical on the mobile page would point to a subpage, where the rel=alternate would point to the primary desktop page, even though the content did not come from that page, necessarily. The primary desktop page is the one that ranks for related keywords. In these cases, the "loop" would be broken. Is this a cause for concern? Could the authority held by the desktop page not be transferred to the mobile version, or the mobile page 'pull away' or disperse the strength of the desktop page if that 'loop' was not connected? Could not setting up the bi-directional tags correctly cause a drop in the organic rankings? -Our developer verified the site is set up according to Google's guidelines for identifying device screen size and serving appropriate version of page. -Are there any tools or utilities that I can use to identify issues, and/or verify everything is configured correctly? -Are we missing anything important in the set-up/configuration? -Could the use of a brand new subdomain 'm.' in and of itself be causing issues? -Have I identified any negative seo practices or pitfalls? Am I missing or overlooking something? While i would have preferred maintaining a single, responsive, site with mobile css, it was not realistic given the various layouts, and owner's desire to only offer the top pages in mobile format. The mobile site may have nothing to do with the organic drop, but I'd like to rule it out if so, and I have so many questions. If anyone could address my concerns, it would be greatly appreciated. Thanks! Greg
Web Design | | seagreen0 -
What seo benefit does setting up a photo gallery where each photo is a separate web page?
what seo benefit does setting up a photo gallery where each photo is a separate web page? My old SEO guy set up my photo gallery like that claiming that because each photo was a separate page, it added a big seo benefit and i never understood what he was talking about. Maybe alt text on the photo with key phrases in it pointing to my other pages to give my site a theme for google? I'm not really sure. He has since moved away and i am considering redoing the photo gallery to multiple images on one page to be more user friendly to my users. This photo gallery is 3 years old and the photos might have some page rank to them helping my site so i don't want to remove this gallery if there really is a benefit to it and it will hurt my site. I once removed four static page rank 3 pages from my site that weren't used for my site anymore and my rankings dropped 5 positions. Thoughts anyone? Thanks! Ron
Web Design | | Ron100 -
Unable to set preferred domain, can I verify a site that's already redirected?
I'm in the process of trying to set a preferred domain in webmaster tools -- to set our www version as preferred vs. the non www. version. IT is already redirecting non-www to www, but I get this message when trying to change settings "Part of the process of setting a preferred domain is to verify that you own http://mnn.com/. Please verify http://mnn.com/." While we own the domain, I am not sure how we can have Google access a file at [http://mnn.com/some_file when we are forwarding all requests for non-www to our www site.
Web Design | | Aggie
Note: The apache rewrite predates me and I'm not sure how / why we have two domains set up, but I'm trying to fix the preferred domain now.Am I able to verify the non version once the redirect is in place.Any ideas??? Help???Thanks!Lisa0 -
Getting a lot more duplicate content warnings than I expected.
I run WordPress on many of my sites and a site crawl has found MANY duplicate content pages on the latest domain I started a campaign for. I expected to see quite a lot on the tag pages that only had one post but even tag pages with multiple posts and author and category pages with many posts are showing as duplicate content. Is this normal for a WordPress site to have so much duplicate content warnings from the taxonomy pages? I have the option to bulk noindex, follow the category and tag pages but should I do it? I get some traffic directly to the tag pages so removing the pages from search results would dent the traffic of the site a little (generally high bounce rate, low engagement traffic anyway) but could removing the apparent duplicate content actually improve the article pages themselves? Or does anyone have any WordPress specific advice for making the pages not duplicate content? I've toyed with the idea of just displaying excerpts but creating manual excerpts for the 4 years worth of posts, some of which I have no personal knowledge of the subject matter so other suggestions are welcome.
Web Design | | williampatton0 -
Hi Everybody. I have a large site that is made up of the main site then a large support site. The support site has a lot of overlapping content and similar titles. Would it be beneficial to separate the two? Thank you. All answers appreciated.
Hi Everybody. I have a large site that is made up of the main site then a large support site. The support site has a lot of overlapping content and similar titles. Would it be beneficial to separate the two? Thank you. All answers appreciated.
Web Design | | arithon0 -
Best way to set up a site with multiple brick and mortor locations across Canada
I have a client who is expanding his business locations from 2 cities to 3, and working towards having 10+ locations across Canada. Right now we're building location based landing pages for each city, as well as keyword targeted landing pages for each city. For example, landing pages for "Vancouver whatever clinic" and "Calgary whatever clinic" as well as for "Vancouver specific service", and "Calgary specific service". This means a lot of landing pages will need to be created to target each of 10 or so desirable "service" keywords for each city's location. I've no issue with this, however I was wondering how other companies go about this? What's the best way to be relevant for certain "service" based keyword searches in each city? Many of the "service" keywords are 'localized' meaning they will show Google Places results for local brick and mortar businesses for each location. I'm quite good at optimizing locally for this type of thing. However, many of the "service" keywords are not yet 'localized' by Google, I'd want to have my client webpages show well in the SERP's. for these 'non-localized' "service keywords" as well. the new site will be built in WordPress
Web Design | | AndyKuiper0