The SPE API constructor (java in this example) uses a single miliseconds parameter as the read_timeout (in miliseconds). This parameter is used when trying to connect to a server AND when scanning a file.
//Java example
ScanEngine.createScanEngine(scanEngineInfoVector, READ_TIMEOUT_MILISECONDS, LOAD_BALANCING_HOST_OFFLINE_DURATION)
If the file is a large file, the scan will take a long time, but if a server is not available, it should not sit waiting for the same long duration(it should know quickly). For this reason there should really be two timeout parameters (scan_timeout and connect_timeout) rather than the single parameter READ_TIMEOUT_MILISECONDS.