Are you running a Connector Framework Server, and seeing multiple kvoop processes? Are you wondering what these are, and whether seeing so many of them is a problem?

The kvoop process is the KeyView Out-of-Process module. It provides CFS with functionality to handle files sent to it, including:

• File format identification
• Meta-data extraction
• Sub-file extraction
• Text filtering
• Character encoding conversion

Basically, kvoop provides the core import functionality for text-based and binary documents with encoded text. Performing these tasks outside of the main CFS process improves flexibility, stability, and resource usage. A KeyView session will typically make use of one kvoop for identification, and another for more in-depth processing.

Are you seeing more than two kvoops running? Check your CFS configuration file for the ThreadCount parameter under [ImportService]. CFS will start a KeyView session for every import thread used, so you will typically wind up with twice this number of kvoops. These sessions will persist while CFS is running, so that it is ready to handle documents as soon as they come in. (Individual kvoops may be restarted automatically, on a periodic basis.)

If the kvoop count still seems high, divide the number of processes by your ThreadCount value. Chances are, this will divide evenly. Some Import Tasks and lua functions may invoke additional KeyView sessions to process files, and these will also be loaded on a per-thread basis. If you’re seeing kvoop processes using more resources than you can afford to allocate to CFS, you may need to decrease the number of threads you’re using, or check whether these additional import tasks are truly needed.

If you’re still seeing what you think are unusual numbers of kvoop processes, or they seem to be using a lot of resources when there shouldn’t be any significant processing going on, take a look at the CFS import log. It may help identify errors that are occurring or documents that are taking longer to process than usual.