Hello, I'm working through using libgfapi and am curious about the number of background threads started. I have a simple cluster with four bricks on four peers. On my client machine I run a program that does this:
glfs_new()
glfs_set_volfile_server()
glfs_set_logging()
glfs_init()
pause()
glfs_fini()
pause()
At the first pause, right after glfs_init, there are four child threads created. Is it a coincidence that the api creates the same number of threads as cluster peers? I'd hate to think that if I had a thousand peers I'd have a thousand threads.
Second, the threads don't shut down after the call to glfs_fini().
Can anyone shed some light on what's going on?