I'm looking for a web based intranet search solution to index some intranet network shares with PDF / Doc / Textfiles etc. and maybe also an intranet wiki. Microsoft has the Search Server Express, which looks promising but the minimum requirements are far too high for my needs. I would prefer a small footprint solution, which could be run as VM like the turnkey linux appliances or something similar.

The ideal would be something like Google Search or Copernic Search running as web server standalone, although with some authentication provided by the SBS in our network.

This answer looked quite promising, but I would prefer something more easier to handle/install, where you mainly have to configure the places to index and maybe some document types and exclusions.

Does anyone has a recommendation here or some pros / cons?

After some time experimenting with Constellio I found it rather hard to install. That is maybe also a result of my lack of experience with Apache Tomcat, but it didn't really ran out of the box or even when I got it to run, only the web indexer worked, other things silently failed. It has a nice UI, but it would be really helpful, if the were any preinstalled virtual applicances to run out of the box as a starting point. So I'm still looking for an alternative ( or a running Constellio installation). Documentation is also relatively small.


Like a Google Mini? (didn't think they still had those, but it looks that way)

  • Well, yes but with no extra hardware. I mean, there are ways to have the built in windows search available in the intranet, but mainly only for Win7 clients, not XP. It would be completely sufficient, if I could have something like the Windows7 search available on all Desktops and centrally managed from the server. I wonder, why there are so few ready-to-use VM based appliances out there...it doesn't looks so complicated to me. – Florian Storck Oct 6 '11 at 10:26
  • some addition: of course any open source based search engine is sufficient, as long it supports indexing PDFs and Docs... web crawling would be nice, but crawling network shares is completely sufficient. – Florian Storck Oct 6 '11 at 10:28

You could have a look at Constellio, I never deployed it but I tried it on a test VM and it was attractive. Give it a try it's opensource (with pro support if needed).

  • looks interesting... – Florian Storck Oct 6 '11 at 11:46
  • hmm...4GB recommened, 2GB minimum...that's quite a lot... – Florian Storck Oct 6 '11 at 12:39
  • Java... as always, from my experience it consumes much more ram than other languages. Some will say that it's due to bad implementation but if this is true I never saw a good one. – Shadok Oct 6 '11 at 13:54
  • I just wonder why all indexing solutions seem to need really big memory. I know, if you want great performance you keep as much as possible in memory, but we're a small business, most of the time, search will be idle...so it's no problem if it would take a few seconds more to search... – Florian Storck Oct 7 '11 at 7:40
  • A few seconds can make a big difference from the point of view of a user. Many of them will not wait and go search the document themselves rather than waiting, the brain hates stalling. – Shadok Oct 10 '11 at 14:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.