Coda file system
Download as PPT, PDF24 likes29,778 views
Coda is a distributed file system developed at Carnegie Mellon University, focusing on scalability, constant data availability, and replication. It utilizes a hierarchical namespace and unique file identifiers (fid) for files and directories, allowing caching and disconnected operation. Coda's architecture includes dedicated file servers (vice nodes) and client machines (virtue nodes), with a robust security system for access control and a reliable RPC system for communication.
![
Versioning vector (Coda Version Vector) when partition happens: [1,1,1]
Client A updates file versioning vector in its partition: [2,2,1]
Client B updates file versioning vector in its partition: [1,1,2]
Partition repaired compare versioning vectors: conflict!](https://image.slidesharecdn.com/codafilesystem-131101080606-phpapp01/85/Coda-file-system-16-320.jpg)
Coda file system
- 2. • • • • • • • • • Introduction to Coda File System Naming and Location Architecture Caching and Replication Synchronization Communication Fault Tolerance Security Summary
- 3. • Coda (constant data availability) is a distributed file system that was developed as a research project at Carnegie Mellon University in 1987 under the direction of Mahadev Satyanarayan. • Coda’s design goals: • Scalability • Constant data availability • Transparency • Security • Consistency
- 4. • The name space in Coda is hierarchically structured as in UNIX and is partitioned into disjoint volumes. • A volume consists of a set of files and directories located on one server, and is the unit of replication in Coda. • Each file and directory is identified by a 96-bit-long unique file identifier (FID) . Replicas of a file have the same FID.
- 5. • An FID has 2 components: 1. A 32-bit RVID (Replication Volume Identifier) of the logical volume that the file is part of. 2. A 64-bit file handle, i.e. vnode, that uniquely identifies the file within a volume.
- 6. Each file in Coda belongs to exactly one volume Volume may be replicated across several servers Multiple logical (replicated) volumes map to the same physical volume
- 8. It works by implementing the following functionalities : 1. Availability of files by replicating a file volume across many servers 2. Disconnected mode of operation by caching files at the client machine
- 9. Coda File System is divided into two types of nodes: 1. Vice nodes: dedicated file servers 2. Virtue nodes: client machines
- 10. The internal organization of a Virtue workstation: is designed to allow access to files even if server is unavailable and uses Virtual File System to intercept calls from client application
- 11. Coda uses RPC2: a sophisticated reliable RPC system Start a new thread for each request, server periodically informs client it is still working on the request
- 12. • • Coda servers allow clients to cache whole files Modifications by other clients are notified through invalidation messages which require multicast RPC a) Sending an invalidation message one at a time b) Sending invalidation messages in parallel
- 13. Session A Client Server Open (RD) Invalidate File f Open (WR) File f Close Close Client Time Session B
- 14. Session A Session C Client A Open (RD) File f Server Open (WR) Invalidate (Callback Break) Close Close File f OK(no file transfer) File f Close Client B Open (RD) Open (WR) Close Time Session B • Scalability • Fault Tolerance Session D
- 15. Data structures: • VSG (Volume Storage Group): • Set of servers storing replicas of a volume • AVSG (Accessible Volume Storage Group): • Set of servers accessible to a client for every volume the client has cached
- 16. Versioning vector (Coda Version Vector) when partition happens: [1,1,1] Client A updates file versioning vector in its partition: [2,2,1] Client B updates file versioning vector in its partition: [1,1,2] Partition repaired compare versioning vectors: conflict!
- 17. HOARDING: File cache in advance with all files that will be accessed when disconnected EMULATION: when disconnected, behavior of server emulated at client REINTEGRATION: transfer updates to server; resolves conflicts
- 18. • • Hoard database Cache equilibrium: • There is no uncached file with a higher priority than any cached file. • The cache is full, or no uncached file has nonzero priority. • Each cached file is a copy of the one maintained in the client’s AVSG. • Hoard walk
- 19. Coda’s security architecture consists of two parts: •The first part deals with setting up a secure channel between a client and a server using secure RPC and system-level authentication. •The second part deals with controlling access to files.
- 21. Operation Description Read Read any file in the directory Write Modify any file in the directory Lookup Look up the status of any file Insert Add a new file to the directory Delete Delete an existing file Administer Modify the ACL of the directory
- 22. • Peter J. Braam, The Coda File System, www.coda.cs.cmu.edu.
Editor's Notes
- #4: Security Secure channels Access control on directories Client Caching Callback promise Callback break Scalability Whole file caching Smart/unreliable clients, dumb/reliable servers No system wide election Disconnected operation Optimistic update strategy and basic conflict resolution Allow files to be edited from local cache Stateful Venus Cache
- #8: Clients have access to a single shared name space. Notice Client A and Client B!
- #17: Unit of replication: volume Volume Storage Group (VSG): set of servers that have a copy of a volume Accessible Volume Storage Group (AVSG): set of servers in VSG that the client can contact Use vector versioning One entry for each server in VSG When file updated, corresponding version in AVSG is updated
- #18: Coda has been designed for high availability, which is mainly reflected by its sophisticated support for client-side caching and its support for server replication. An interesting aspect of Coda that needs further explanation is how a client can continue to operate while being disconnected, even if disconnection lasts for hours or days. HOARDING normal state of a client Client connected to (at least) one server that contains a copy of the volume. While in this state, the client can contact the server and issue file requests to perform its work, simultaneously, it will also attempt to keep its cache filled with useful data EMULATION AVSG=0 behavior of a server for the volume is emulated on the client’s machine. This means that all file requests will be directly serviced using the locally cached copy of the file. Note that while a client is in its EMULATION state, it may still be able to contact servers that manage other volumes. In such cases, disconnection will generally have been caused by a server failure rather than that the client has been disconnected from the network. REINTEGRATION when reconnection occurs, the client enters the REINTEGRATION state in which it transfers updates to the server in order to make them permanent. It is during reintegration that conflicts are detected and, where possible, automati-cally resolved. As shown in the figure, it is possible that during reintegration the connection with the server is lost again, bringing the client back into the EMULATION state