It is possible to run multiple instances of the Hurd
in parallel, on a single instance of Mach.  Other than
performance crosstalk, they are essentially isolated.
Practically, as many devices do not allow multiple
non-cooperating users, e.g., hard drive and network
this is not currently possible.  It can be overcome,
however, by virtualizing these problematic devices.

When extra hardware is not available, it is possible to
use a sub-hurd.  A sub-Hurd is like a neighbor Hurd,
however, makes use of some resources provided by another
Hurd.  For instance, backing store and the console.
([[SubhurdHowto]])

Sub-hurds are extremely useful for debugging core
servers as it is possible to attach to them with gdb
from the parent ([[DebuggingViaSubhurds]].  This avoids
deadlock, e.g., when the instance of gdb stops the
server but requires its use.
(Note: it is possible to use [[NonInvasiveDebugging]],
but this is less flexible.)