404 page not found after site migration

[PASSLtd]

Hi,

A question from our developer.

We have an issue in Google Webmaster Tools.

A few months ago we killed off one of our e-commerce sites and set up another to replace it. The new site uses different software on a different domain. I set up a mass 301 redirect that would redirect any URLs to the new domain, so domain-one.com/product would redirect to domain-two.com/product. As it turns out, the new site doesn’t use the same URLs for products as the old one did, so I deleted the mass 301 redirect.

We’re getting a lot of URLs showing up as 404 not found in Webmaster tools. These URLs used to exist on the old site and be linked to from the old sitemap. Even URLs that are showing up as 404 recently say that they are linked to in the old sitemap. The old sitemap no longer exists and has been returning a 404 error for some time now. Normally I would set up 301 redirects for each one and mark them as fixed, but there are almost quarter of a million URLs that are returning 404 errors, and rising.

I’m sure there are some genuine problems that need sorting out in that list, but I just can’t see them under the mass of errors for pages that have been redirected from the old site. Because of this, I’m reluctant to set up a robots file that disallows all of the 404 URLs.

The old site is no longer in the index. Searching google for site:domain-one.com returns no results.

Ideally, I’d like anything that was linked from the old sitemap to be removed from webmaster tools and for Google to stop attempting to crawl those pages.

Thanks in advance.

[CleverPhD]

I agree that the 301 redirect would be your best option as you can pass along not only users but the bots to the right page..  You may need to get a developer in to write some regular expressions to parse the incoming request and then automatically find the correct new URL.  I have worked on sites with a large number of pages and using some sort of automation is the only way to go.

That said, if you simply want to kill the old URLs you can show the 404s or 410s.  As you mention, then you end up with a bunch of 404 errors in GWT.   I have been there too, it's like damned if you do, damned if you don't.  We had some URLs that were tracking URLs from an old site and we are now here a year later (been showing 410s for over a year on the old tracking URLs) they still show up in GWT as errors.

We are trying a new solution for how to remove these URLs from the index without getting 404 errors.  We show a 200 and then we put up a minimal html page with the meta robots noindex tag.

http://support.google.com/webmasters/bin/answer.py?hl=en&answer=93710

"When we see the noindex meta tag on a page, Google will completely drop the page from our search results, even if other pages link to it. "

So, we allow Google to find the page, get a 200 (so no 404 errors), but then use the meta noindex tag to tell Google to remove it from the index and stop crawling the page.

Remember, this is the "nuclear" option.  You only want to do this to remove the pages from the Google index.  Someone mentioned using GWT to remove URLs, but if I remember correctly, you only have so many pages you can do this with at a time.

If you list the files within the robots.txt. Google will not spider the files, but then if you remove the page from robots.txt file, they will start to try spidering again.  I have seen Google come back a year later on URLs when I take them out of robots.  This is what happened to us and so we tried just showing the 410/404, but Google still keeps crawling.   We recently moved to this option with the 200/noindexmeta and it seems to be working.

Good luck!

[Highland]

You can but the 404s should stop being crawled on their own. There's a webmaster tool that you can use to make that happen faster as well

http://support.google.com/webmasters/bin/answer.py?hl=en&answer=64033

[PASSLtd]

Yeah it's a 404 http://www.tester.co.uk/17th-edition-equipment/multifunction-testers/fluke-1651b-multifunction-installation-tester

with over 200,000 404's its a lot to go through and 301. For some reason they it got migrated they just pointed the old url to a new one replacing the root domain name without creating matching url's. Doh.

I was thinking about robot.txt filling them all?

[Highland]

A 404 should cause Google to de-index the content. Go to one of the bad URLs and view the headers to make sure that your webserver is returning a status 404 and not just a 404 "page".

As hard and time consuming as it might be, I would still pursue a 301 option. It's the cleanest way to resolve the issue. Just start nibbling at it and you can make a dent. Doing nothing just lets the problem grow.