Knowledgebase

SpaceObServer is compatible with all current versions of the following databases: MS SQL Server, MSSQL Azure, MySQL and Firebird. Please find more details here. You can also use the free Express editions of the commercial SQL servers.

Since Windows 2008 Server and Windows Vista, Microsoft uses 'Client Caches' in order to speed up the queries about gathering file and directory meta information on network drives. This approach also helps to utilize the available network bandwidth more efficiently. Anyway, these caches always return an incorrect value for the NTFS hardlinks (always shown as '1').

Caches can be disabled by the following steps:
1. Run 'regedit'
2. Open the path 'HKEY_LOCAL_MACHINE\System\CurrentControlSet\services\Lanmanworkstation\Parameters'.
3. Choose "New > DWORD value" in the right click menu and set the name to 'FileInfoCacheLifetime' and the value to '0'.
4. Save the registry and quit 'regedit'.

After these steps SpaceObServer should work properly with hardlinks. For further information about the subject please visit the Microsoft Website.
Note: In case a file has been recently opened, the number of hardlinks will be shown as 1. After a few seconds the value will be shown correctly again.

Since DFS shares do not represent a physical directory structure, the option 'Follow mount points and symbolic links' must be activated in the "Configure Scans" dialog. By default, folders that are not physically located at the location to be scanned are excluded from the scan.

In order to increase the scan performance of SpaceObServer, please consider the following:

1. Do not use the 'SpaceObServer Embedded' database for data storage. A SQL Server like MS SQL Server or My SQL provides a much better performance for scanning and exporting data in combination with SpaceObServer.

2. Use a well performing database server, with an up-to-date CPU. The database is the bottleneck of the scan performance. 70-80% of the scan performance and the workload is hold by the database. Please check if the machine on that the database server is running has a high CPU load. If so, consider transferring it to a different machine. Furthermore, a database server will greatly benefit from additional free RAM, so extending the RAM usually increases the performance of the database server

3. If you are scanning network shares, another reason for a slow scan can be a slow network connection to the target machine. Local scans run faster than network scans. So if it is possible we recommend installing SpaceObServer directly on the file server and accessing the scanned information with SpaceObServer Remote Clients or the SpaceObserver WebAccess. If you however prefer network scans, we do not recommend to scan more than 3-5 machines with one instance of SpaceObServer, but this depends totally on the number of files and folders on the servers as well as it depends on the update rate of the files and folders on the servers.

4. The option 'Track NTFS Alternate Data Streams and NTFS Hard Links' will also decrease the scan speed, if activated. This option can be toggled in the  'Configure Scans' dialog. Errors during the scan (database timeouts, missing access permissions etc.) can also cause slow scans. All errors are logged in your Windows Event Log (Category 'Application') under the source 'SpaceObServer Service' and can be viewed there.

5. 'The more entries in the database tables, the slower the scans'. This is why multiple scans are split to multiple database tables.
In addition, you could split the server scans to multiple scans of UNC shares ('\\Server1\share1\'; '\\Server1\share2\'; ). This would result in smaller tables and faster scans.

(6. The index structure of the SpaceObServer database is optimized to do both, execute scans and access the information from the user interface. By customizing the indices of SpaceObServer you may optimize the performance. We do not recommend doing this, since if you manipulate the existing index structure this could have huge consequences for the performance of the scans and the user interface.)

The first scan of SpaceObServer always takes longer than the subsequent scans , since the whole database structure has to be build and the complete file system information is collected during the initial scan. Update scans only collect the file system changes since the last scan. The speed of update scans basically depends on the number of file system changes, the number of changed files and the database size (amount of data that has already been collected into the database).

Yes, multiple scans can be added via the OLE Automation of SpaceObServer. Further information can be found here.

To make sure that the SpaceObServer Remote Client does not connect to the remote file server, go to "Tools > Options > View > General" and activate "Determine Icons based on file extension only", deactivate "Show files deleted since last scan grayed" and deactivate "Show long tool tips in the directory tree".

The SpaceObServer Add-on SpaceObServer Web Access can access scan results via your internet browser and provides only reporting functionality. You can also do this by using the SpaceObServer Remote Client, which is a separate product. Just create a user profile on your SQL server without write access to the SpaceObServer database. Ensure that the connection of the SpaceObServer Remote Client uses this user profile. The Remote Client will be able to create reports for any scanned shares of the database, but will not be able to schedule scans. All database manipulating features of the user interface will be deactivated. Many of our customers are already providing the data collected by SpaceObServer to their users this way.

Beside of a simple deletion of redundant duplicate files SpaceObServer offers the possibility to deduplicate chosen files. This will recover the disk space allocated by your duplicate files but leave links to the files content in your file system.
Note: Before deleting or deduplicating any duplicate files you should ensure that the listed files are really redundant and not used, e.g. as part of a backup.

By default the "Duplicate Files" list shows all files with same name, size and last change date. You can optionally use MD5 checksums to detect files with equal content. In this case you have to activate the option "Calculate MD5 checksums for files" (See "Configure Scans"), which results in slightly higher allocated database space and some increased performance usage of the "SpaceObServer Service".You can customize the combination of these criteria used for identifying duplicates in the options. Use the function "Compare Files Byte by Byte" from the context menu of this list to check the content of the selected files byte by byte.