diff --git a/gossip/README.md b/gossip/README.md index 7ab4e41..02aafe3 100644 --- a/gossip/README.md +++ b/gossip/README.md @@ -1,7 +1,7 @@ ## Gossip If you only read one of these papers, start with the classic Demers, et al paper: -* [Epidemic Algorithms for Replicated Database Maintenance] (http://bitsavers.informatik.uni-stuttgart.de/pdf/xerox/parc/techReports/CSL-89-1_Epidemic_Algorithms_for_Replicated_Database_Maintenance.pdf) +* [Epidemic Algorithms for Replicated Database Maintenance](http://bitsavers.informatik.uni-stuttgart.de/pdf/xerox/parc/techReports/CSL-89-1_Epidemic_Algorithms_for_Replicated_Database_Maintenance.pdf) # Peer sampling services Briefly, a peer sampling service is a system that maintains a restricted set (partial view) of the all machines participating in a gossip system.