locked
Site unavailable from dedicated index server RRS feed

  • Question

  • I had a server setup consisting of one WFE running search, web program, excel calculations, etc. (multi-function) with a dedicated database server. This setup has worked fine so far.

    However, I want to add a dedicated indexing server to the farm. After some hiccups with the initial installation (due to some version mismatches, now patched and fixed) the server was installed and I could start the 'Office SharePoint Server Search' service, set it up to index and be a dedicated index server (ie. use only that server, not all WFE's). The name of the new server is INDEX, while the old server was simply called MOSS. I changed the SSP to use INDEX as the new indexing server.

    However, in the SSP Search Administration I get errors that it cannot access INDEX. I tried accessing http://site/ from within the indexing server and I couldn't authenticate (tried entering valid credentials and got 401). After some searching around I disabled the loopback check. I can now access the site, but I get presented with a standard Sharepoint "Runtime error" page. I tried setting "customErrors='Off'" in the web.config, but it does not change anything - since I don't want to start showing errors on the "public" WFE, I haven't changed the web.config on MOSS yet.

    I need some pointers as to what could be wrong. The new server is running, all services are up and running, I get no errors when restarting any of them, changing the index server on the SSP or anything like that - it's only when I actually try to view the site that I get the errors.

    FYI: It's a windows server 2003 setup using MOSS 2007 version 12.0.0.6504 (both MOSS and INDEX).

    Friday, February 25, 2011 12:07 PM

Answers

  • I would suggest you enable the Web application role for the INDEX server. In such a way, you can troubleshoot the issue with the web.config on the INDEX server.

    Once you enable the web application role, change the host entries for the INDEX server to point to itself. The other advantages with it is, when crawl would run, it would hit itself without impacting the end user on the other WFEs.

    Thanks,
    Dip


    Dip
    http://dzeee.net/sharepoint
    • Marked as answer by Leoyi Sun Thursday, March 3, 2011 2:13 AM
    Saturday, February 26, 2011 4:03 PM