Distributed File System | prof. Giovanni Chiola | ![]() |
Files should be CONTINUOSLY accessible all the time (even in case of transient failures). | ||
Client failures can be overcome by fast substitution. | ||
Complete network failures can only be overcome by local caching on the client machine (detached operation). | ||
Partial network or server failures can be overcome by redundancy and redirection of requests to currently available servers. | ||
| ||
previous | intro04.htm | next |