Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Automatic Restart of a Stopped Process by a Managing Process

IP.com Disclosure Number: IPCOM000103816D
Original Publication Date: 1993-Feb-01
Included in the Prior Art Database: 2005-Mar-18
Document File: 2 page(s) / 38K

Publishing Venue

IBM

Related People

Carlsen, K: AUTHOR [+7]

Abstract

Disclosed is a method for a tending daemon to restart sub-processes that have terminated prematurely.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 85% of the total text.

Automatic Restart of a Stopped Process by a Managing Process

      Disclosed is a method for a tending daemon to restart
sub-processes that have terminated prematurely.

      In a distributed client/server based file backup and restore
environment, different services are provided by individual daemons
that handle requests made by client operations.  On a given node in
the system, one or more daemons may be running in the background,
providing services such as message handling and inventory database
transactions to all other nodes in the system.  To assure that the
daemons on a given node are maintained and to provide a common method
for controlling all such services via a remote procedure call
interface, a tending daemon is put in place to handle startup,
shutdown, and error recovery/restart of all other services.

      An important function of the tending daemon is to restart any
services that it is tending which terminate prematurely because of a
computer shutdown, accidental termination by a privileged user, or
other fault.  Under the UNIX* environment, the tender process
receives a "CHILD DIED" signal when any sub-process it creates
terminates.  The tender handles this signal by logging the failure
and attempting to restart the failed service.  Each service is
designed so that it goes into error recovery, if needed, whenever it
is started.  Therefore, a successful restart of the service is all
that is needed to initiate error recovery and restore the...