Software Tools and Libraries for Fault Tolerance

Yennun Huang, Chandra Kintala and Yi-Min Wang
AT&T Bell Laboratories
Murray Hill, NJ 07974

Abstract:

Software fault tolerance is the task of detecting and recovering from
faults that are not handled in the underlying hardware or operating system
layers of an application.  We describe four reusable software components,
watchd, libft, libckp and REPL, that perform automatic detection and
restart of failed processes, checkpointing and recovery of in-memory data,
and replication and synchronization of files in an application.
These components have been ported to a number of UNIX platforms and can
be used in any application with minimal programming effort.  Several
telecommunications products in AT&T have enhanced their fault-tolerance
capability using these components.  Experience with those products to date
indicates that these modules provide efficient and economical means to
increase the level of fault tolerance in an application.