#Number
TR-PDS-1995-013

#Title
Sharing state information for fast recovery in a $k$ faulty system

#Author
J. Roger Mitchell

#Abstract
There have largely been two approaches to recovery in a distributed system
following a failure: checkpointing to stable storage and replication
of processes.  Both allow a process to fail and recover assuming
stable storage or a replicate is not adversely affected during the
failure.  In the approach presented here, we use checkpointing at
another process which does not require either replicates or stable
storage.  This provides the benefits of lower communication overhead
as compared to replicates and faster recovery as compared to stable
storage.  

#Bib
@InProceedings{,
  author = 	 "J. Roger Mitchell",
  title = 	 "Sharing state information for fast recovery in a $k$ faulty system",
  booktitle =	 "the publication name",
  address =	 "location of publication",
  month =	 "",
  note =	 1995,
  note =	 "available via ftp or WWW at maple.ece.utexas.edu
		  as technical report TR-PDS-1995-013"
}