How does the NFSNET client communicate with the server?
The client communicates with the server using an HTTP based protocol. This allows us to provide executables for most environments that have basic Internet connectivity.

How often does the client communicate with the server?
The frequency varies from project to project however, when we choose the parameters for a particular factoring project we try to select them such that one work assignment completes within 1 to 5 hours.

How is the data transferred?
As the siever runs, the data accumulates in a file called results.txt which resides in processors/p0 directory. Once a sieve run is complete the file is then uploaded to the server using the same HTTP protocol that is used getting assignments.

How much data is transferred?
This also varies from project to project but it also varies with the particular work unit that is assigned. Currently, NFSNET uses what is called a line siever. The nature of line sieving is such that assignments in the lower end of the region to be sieved will produce more relations than those in the higher end. In general, the size of an individual results file will vary from as little as 10K to as much as 500K.

What if my network connection is down?
If the NFSNET client can't communicate with the server it will pause and retry again after a while.

All FAQ sections:
    System Requirements
    Running the NFSNET Client
    Communications
    About the Number Field Sieve