UNIX System Administration Handbook - Evi Nemeth [493]
See page 709 for more information about lpd.
lpsched: manage ATT printing
lpsched is the ATT version of the line printer daemon. It receives print jobs from the lp program and queues them for printing. When an appropriate device becomes available, lpsched forks a process to manage the actual printing.
See page 721 for more information about lpsched.
rlpdaemon: print from BSD to HP-UX
rlpdaemon is an HP-UX daemon that allows HP’s SysV-ish lpsched system to accept print requests from BSD-style systems. See page 732 for more information about rlpdaemon.
28.6 NFS DAEMONS
The following daemons are part of the NFS file sharing system. We give only a brief description of their functions here; Chapter 17 describes them in detail.
nfsd: serve files
nfsd runs on file servers and handles requests from NFS clients. On some systems, this daemon is called rpc.nfsd.5
In most NFS implementations, nfsd is really just a part of the kernel that has been dressed up as a process for scheduling reasons. nfsd takes a single argument that specifies how many copies of itself to fork. Some voodoo is involved in picking the correct number of copies; see page 499 for specifics.
mountd: respond to mount requests
mountd (which is called rpc.mountd on some systems) accepts filesystem mount requests from potential NFS clients. It verifies that each client has permission to mount the requested directories. mountd consults an export list to determine which applicants are legitimate.
amd and automount: mount filesystems on demand
amd and automount are NFS automounters, daemons that wait until a process attempts to use a filesystem before they actually mount it. The automounters later unmount the filesystems if they have not been accessed in a specified period of time.
The use of automounters is very helpful in large environments where dozens or hundreds of filesystems are shared on the network. Automounters increase the stability of the network and reduce configuration complexity, since all systems on the network can share the same amd or automountd configuration. We cover the use of these daemons in detail, starting on page 504.
lockd and statd: manage NFS locks
Although lockd and statd (aka rpc.lockd and rpc.statd) are two distinct daemons, they always run as a team. lockd maintains advisory locks (a la flock and lockf) on NFS files. statd allows processes to monitor the status of other machines that are running NFS. lockd uses statd to decide when to attempt to communicate with a remote machine.
biod: cache NFS blocks
biod (which is known as nfsiod on FreeBSD systems) caches read and write requests on some NFS clients. biod performs both read-ahead and write-behind buffering and greatly improves the performance of NFS.
See page 501 for more information about biod.
28.7 NIS DAEMONS
Several daemons are associated with Sun’s NIS and NIS+ administrative database systems. These systems are described in Chapter 18, Sharing System Files. Note that NIS and NIS+ are distinct and independent systems, despite their similar names. Although both packages originated at Sun, they are now used on many other vendors’ systems as well.
ypbind: locate NIS servers
The ypbind daemon runs on all NIS clients and servers. It finds an NIS server to which queries can be directed. ypbind does not actually process requests itself; it just tells client programs which server to use.
ypserv: NIS server
ypserv runs on all NIS servers. ypserv accepts queries from clients and responds with the requested information. See page 527 for information on how to configure the machines that run ypserv.
ypxfrd: transfer NIS databases
ypxfrd transfers NIS databases to slave servers in an efficient manner. A slave initiates a transfer with the ypxfr command. Whenever a database is changed on the master, it should immediately be pushed