Other File Systems: AFS, Napster

1
Other File SystemsAFS, Napster
2
Recap

• NFS
• Server exposes one or more directories
• Client accesses them by mounting the directories
• Stateless server
• Has problems with cache consistency, locking
  protocol
• Mounting protocol
• Automounting
• P2P File Systems
• PAST, CFS
• Relies on DHTs for routing

3
Andrew File System (AFS)

• Named after Andrew Carnegie and Andrew Mellon
• Transarc Corp. and then IBM took development of
  AFS
• In 2000 IBM made OpenAFS available as open source
• Features
• Uniform name space
• Location independent file sharing
• Client side caching with cache consistency
• Secure authentication via Kerberos
• Server-side caching in form of replicas
• High availability through automatic switchover of
  replicas
• Scalability to span 5000 workstations

4
AFS Overview

• Based on the upload/download model
• Clients download and cache files
• Server keeps track of clients that cache the file
• Clients upload files at end of session
• Whole file caching is central idea behind AFS
• Later amended to block operations
• Simple, effective
• AFS servers are stateful
• Keep track of clients that have cached files
• Recall files that have been modified

5
AFS Details

• Has dedicated server machines
• Clients have partitioned name space
• Local name space and shared name space
• Cluster of dedicated servers (Vice) present
  shared name space
• Clients run Virtue protocol to communicate with
  Vice
• Clients and servers are grouped into clusters
• Clusters connected through the WAN
• Other issues
• Scalability, client mobility, security,
  protection, heterogeneity

6
AFS Shared Name Space

• AFSs storage is arranged in volumes
• Usually associated with files of a particular
  client
• AFS dir entry maps vice files/dirs to a 96-bit
  fid
• Volume number
• Vnode number index into i-node array of a volume
• Uniquifier allows reuse of vnode numbers
• Fids are location transparent
• File movements do not invalidate fids
• Location information kept in volume-location
  database
• Volumes migrated to balance available disk space,
  utilization
• Volume movement is atomic operation aborted on
  server crash

7
AFS Operations and Consistency

• AFS caches entire files from servers
• Client interacts with servers only during open
  and close
• OS on client intercepts calls, and passes it to
  Venus
• Venus is a client process that caches files from
  servers
• Venus contacts Vice only on open and close
• Does not contact if file is already in the cache,
  and not invalidated
• Reads and writes bypass Venus
• Works due to callback
• Server updates state to record caching
• Server notifies client before allowing another
  client to modify
• Clients lose their callback when someone writes
  the file
• Venus caches dirs and symbolic links for path
  translation

8
AFS Implementation

• Client cache is a local directory on UNIX FS
• Venus and server processes access file directly
  by UNIX i-node
• Venus has 2 caches, one for status one for data
• Uses LRU to keep them bounded in size

9
Napster

• Flat FS single-level FS with no hierarchy
• Multiple files can have the same name
• All storage done at edges
• Hosts export set of files stored locally
• Host is registered with centralized directory
• Uses keepalive messages to check for connectivity
• Centralized directory notified of file names
  exported by the host
• File lookup client sends request to central
  directory
• Directory server sends 100 files matching the
  request to client
• Client pings each host, computes RTT and displays
  results
• Client transfers files from the closest host
• File transfers are peer-to-peer central
  directory not part

10
Napster Architecture
Napster Directory Server 1
H1
Napster Directory Server 2
H2
IP Sprayer/ Redirector
Firewall
Network
Napster Directory Server 3
Napster.com
H3
11
Napster Protocol
Napster Directory Server 1
H1
I have metallica / enter sandman
Napster Directory Server 2
H2
IP Sprayer/ Redirector
Network
Firewall
Napster Directory Server 3
Napster.com
H3
12
Napster Protocol
Napster Directory Server 1
H1
I have metallica / enter sandman
Napster Directory Server 2
H2
IP Sprayer/ Redirector
Network
Firewall
who has metallica ?
check H1, H2
Napster Directory Server 3
Napster.com
H3
13
Napster Protocol
Napster Directory Server 1
H1
I have metallica / enter sandman
Napster Directory Server 2
H2
IP Sprayer/ Redirector
Network
ping
ping
Firewall
who has metallica ?
check H1, H2
Napster Directory Server 3
Napster.com
H3
14
Napster Protocol
Napster Directory Server 1
H1
I have metallica / enter sandman
Napster Directory Server 2
H2
IP Sprayer/ Redirector
Network
ping
ping
Firewall
who has metallica ?
check H1, H2
Napster Directory Server 3
transfer
Napster.com
H3
15
Napster Discussion

• Issues
• Centralized file location directory
• Load balancing
• Relies on keepalive messages
• Scalability an issue!
• Success ability to create and foster an online
  community
• Built in ethics
• Built in faults
• Communication medium
• Had around 640000 users in November 2000!

16
Other P2P File Systems

• Napster has a central database!
• Removing it will make regulating file transfers
  harder
• Freenet, gnutella, kazaa all are decentralized
• Freenet anonymous, files encrypted
• So not know which files stored locally, which
  file searched
• Kazaa allows parallel downloads
• Torrents for faster download