Testing Yet Another SE

If you’ve been following our internal search engine changes, you saw that we buckled about 2 weeks ago under the load at a peak of 5,000 searches on our site a day.

In the interim, we went with Google’s AdSense for Search services. We like the speed Google can deliver, but the search results are always delayed a bit until Google indexes the newest original articles added each day.

2 weeks worth of search engine testing, multiple long nights of testing by our programming team, multiple index runs one after the other after the other, and lots of tweaking have delivered the new internal search engine that we rolled out yesterday.

This one meets three needs that the previous search engine didn’t: 1) It does not cause a 11-15 hour brownout or blackout while the search engine indexes our site to update the index and 2) it’s fast and 3) it appears robust enough to deal with a high search load.

We are already working on the 2nd and 3rd phases of our internal search engine — so yes, some of the functionality is missing yet but will be back online soon.

Give it a spin: http://www.ezinearticles.com/ and let us know what you think of the search results.

Curious: Should we also provide the Google Search option that we just took down? ie: Do you want to do a search of our site to see what Google has indexed so far? Is that of any value?


Lance Winslow writes:

Indeed, that is an improvement, works very well, I especially enjoyed the highlighting feature to find things fast. I searched; “unmanned aerial vehicle” to find all my UAV articles, very cool. You know this sure helps those who wish to store their writing on your site as a back up so they do not lose them if their PC crashes. Additionally if someone has written a bundle of articles and does not mind sharing them with the world, it would make sense to park them here for safe keeping. I am impressed, hope it goes well for you.

Comment provided November 17, 2005 at 4:24 PM


Merrick writes:

It struck me as odd before that you did not employ a full text search on your database to do the work. Both mysql and postgres allow you to accomplish search without the need to crawl, the index is built in to the database.

Comment provided November 17, 2005 at 4:26 PM


Programmer #1 writes:

We Tried That. It was unable to handle the query load.

Comment provided November 18, 2005 at 1:49 AM


RSS feed for comments on this post.

Leave a comment

Please read our comment policy before commenting.