I just setup ES within our VM Ware system and it was very easy. I got everything configured, tied into our AD structure and enabled several test users. The problem I’m running into is setting up collections. We are a MS 2008R2 AD domain but for some reason when I setup of a collection I keep getting an error.

Error: crawlsmb.c: Error! Could not open directory //fminet.com/fmiitgroup/itdoc: for dir “//fminet.com/fmiitgroup/itdoc” at src/crawlSMB/crawlsmb.c:563

I know this is the correct path because I can map to it and I’m using a Administrator level login. I’ve tried using the user name as @domainname.com and domainname\ before the user name. Is there some type of trust relationship with this new server I need to setup since it’s using Linux? Or is there something else I’m missing. The file path I’m trying to crawl is in the same subnet so that shouldn’t be an issue. I’m been looking through your online documentation but can’t find anything obvious I’m doing wrong. Thanks for any help provided.

And could the issue be the Windows Firewall running on these servers. If so can tell me what ports TCP/UDP need to be opened. I ran the manual scan in EC on the server subnet and it didn’t return any shares, so that got me to thinking maybe it is the firewall.

1) Is “//fminet.com/fmiitgroup/itdoc” the real name of your share, or have you anonymized it?

I see that fminet.com resolve to a website. Is the same server running both your web server and your file server? I tried to find out what server that was, and according to http://web-sniffer.net/ is an Apache web server, which isn’t that common on Windows.

I would recommend that you try using the ip address of the file server to see if that changes anything.

2) Supply the username with the active directory domain name in the format “domain\user”  is normally the correct way.  For example we have a domain called sdtest.local here. My user is “jarleh”. Some file servers accept only jarleh as username, but most needs sdtest\jarleh. (if that works you also have to set domain as user prefix to resolve the username correct at runtime ).

3) If you get stuck try to use smbclient from a Linux box (or from the ES. Just log in as root , the default password will be visible in the console) to debug, because it can give you better error message. smbclient uses the same library as the ES so the parameters that works there should also work on the ES.

> Is there some type of trust relationship with this new server I need to setup since it’s using Linux?
>

No the ES don’t need to be a member of your active directory. However the active directory isn’t really involved when crawling files anyway.

And could the issue be the Windows Firewall running on these servers. If so can tell me what ports TCP/UDP need to be opened.

If it works from windows I would assumed the firewall didn’t block the ES only because it’s on Linux, but I can ask one of our developers about it.

Ok using the IP address worked, along with domainname\username. So it’s crawling now. Now does the initial crawl have to finish before you can see the collection showing up in search?
And I’m confused about the webpages for the users to go to for searching. I see a Public and Private. Everything thing is within our network domain that we will be searching. No one outside our network will be able to search, it’s just for our users. So does that mean we should use Private? And if does this use their AD credentials to login to the Private site? I’ve used it but keep getting a login each time and my AD credentials don’t work. So maybe I’m confusing them.

Ok right now I can only see my collections when setting the permissions to Anonymous. I have added our AD structure and read through the documents where is says:

System permissions, user and group access levels require integration with an existing user system. End users can log in and use the search from the main search page her. Anonymous/public search make the whole collection available for search by everyone, including the whole internet, without having to login. No existing user system is required. The public search page is available her.

So my guestions are:
When setting up System, User and Group access does that use the Private login?
If so when using it I’m prompted for a Username and password. So that should be their normal AD domain user name and password?
If that is true so far each time I’ve tried this it fails and won’t let me in. My user account has been enabled, as has a couple of other accounts we’re using for testing. I’ve tried all various user name possibilities to no avail. Is there something else that needs to be enabled to log in and view the collections?

Ok after playing around and changing various settings I got it to work now on the Private login.

Can you explain the difference on each Collection Access Level Tab the difference between System Permissions and User? We want for Users domain permissions on each of the Shares to come into play so they can’t access files they aren’t suppose to. I’m assuming System Permissions is what I need to choose but I wanted to verify.

Can you explain the difference on each Collection Access Level Tab the difference between System Permissions and User? We want for Users domain permissions on each of the Shares to come into play so they can’t access files they aren’t suppose to. I’m assuming System Permissions is what I need to choose but I wanted to verify.

“System permissions” is normally what you want if you are in a Microsoft Active Directory world with a Windows file server and users shall log in before searching. The system permissions uses information from Microsoft Active Directory and the permission set on file and folders on the file server you have indexed so users will only be able to search files they have access to.

The other access levels (User, Group and Anonymous) are used to overwrite the standard setup if you have special needs. This is normally not what you want for files from a file server.

Ok after playing around and changing various settings I got it to work now on the Private login.

Good! Does that mean that you no longer need answer on the other questions above?

Yes that does answer those questions. I do have some about changing the default template when users login. I did a search and found some answers like:
Clone the default template and edit the main.tpl and head.tpl to change img/common/logo_smaller.gif and img/common/logo_top.jpg .

But I’m not certain how to get our logo files to that server and store them in that location to customize it. I tried mapping with no luck. And do those logo files need to be a certain resolution to fit properly?

And I to would think it would be nice to be able to with a click to change the default template. Hopefully that is on the future roadmap for any updates.

The logo and template can be changed, but there are some limitations. Please see this thread: http://www.searchdaimon.com/forum/viewthread/52/

I’m not familiar with either of those tools. These logo’s are stored on our Windows Servers and would need to copy them some how to the ES server which I’m assuming is Linux based. Are those the only tools you’re aware of that would work or will they work at all for what we need? Can you directly link to those files via a network file path in the template or do they need to be local. We would just need to copy to small JPG and GIF files over so something simple would be best.

Thanks for all your help…

I’m wondering if maybe I could use SSH to connect to transfer those files or maybe FTP. From what I’m reading SSH is disabled and I see the post here about changing that but can’t get logged in as root again no matter what I try. Same with changing the time and date as it’s off by eight hours. But again I need to login as root to do this from what I’m reading. So can you verify the login and password I need to make these changes? Not completely familiar with Linux but I’ve tried su - and a couple of various logins along with the password that it shows on the VMware console screen.

Thanks again…

Ok I was able to get in as root and get the time and date changed. When I run the Date command on the command line it shows the correct date and time now. But when looking in the Search logs in the WEb Console it still is showing the incorrect time of when the Searchs took place, even search that happened after the time zone change. Any ideas on how to get that corrected? Does the web service maybe need to be restarted? If so how do you do this?

Also now that I’m in the root I was able to enable SSH as well, but still can’t seem to log in using Putty. If no longer blocks it but it won’t accept the default password. I can still log in via the VMware console but Putty would be easier.

And then finally I’m trying to find the img/common/ folder to upload our log files but can’t seem to find it. I’ve looked in the Home directory, the root, WWW folders and no luck. Can you give me a more direct path of where it’s located?

Ok I was to get all three of these issues resolved. Now I have another one.

I’m get Insufficient Disk space messages while trying to craw a new collection. It’s at 58646 documents but has stalled. Disk space on drive 2 showed only 5GB used of the 40GB I setup, so it shouldn’t be out of space. I’m also seeing the CPU peak out at 100% that last all night. I finally had to reboot and it’s settled down. But I had given it 8GB of RAM and two CPU’s to work with. I’ve now upped it to 16GB and 4 CPU’s so I can get through the crawls, and now I’m getting the disk space errors. I’ve read several items about this being a known issue. Are there any work arounds? If there is a cache file that is using up space can it be removed? Rebooting doesn’t seem to help.

Good to hear that the crawling issue is resolved. What was it?

Unfortunately so have we seen cases where the ES sees it’s out of space, but there is plenty left. For example her: http://www.searchdaimon.com/forum/viewthread/727/ . Unfortunately no one in Searchdaimon has ever been able to replicate this in any test setups, so we are a little bit lost here.

What kind if VMware systems are you on?

If you run “df -h /boithoData” on the console what does it sees about “Filesystem” and “Avail”?

And then finally I’m trying to find the img/common/ folder to upload our log files but can’t seem to find it. I’ve looked in the Home directory, the root, WWW folders and no luck. Can you give me a more direct path of where it’s located?

All the software is in the /home/boitho/boithoTools folder, so the img/common/ is in /home/boitho/boithoTools/public_html/webclient2/img/common/ .