1 / 15

Outline

Outline. • Introduction (what’s it all about) • Data-centric consistency • Client-centric consistency • Replica management • Consistency protocols. Consistency protocol. Describes the implementation of a specific consistency model. Continuous consistence Sequential consistence

Download Presentation

Outline

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Outline • Introduction (what’s it all about) • Data-centric consistency • Client-centric consistency • Replica management • Consistency protocols

  2. Consistency protocol • Describes the implementation of a specific consistency model. Continuous consistence Sequential consistence Primary-based protocols Replicated-write protocols

  3. Continuous Consistency (1) • Degree of consistency based on deviations of numerical values (due to operation performed)

  4. Continuous consistency

  5. Si propagate a write originated from Sj to Sk. And it meets the above bound \delta_i.

  6. Consistency protocol • Describes the implementation of a specific consistency model. Continuous consistence Sequential consistence Primary-based protocols Replicated-write protocols

  7. Consistency protocol • Primary-based protocols • sequential consistency The result of any execution is the same as if the (read and write) operations by all processes were executed in some sequential order specified by its program • Data item x has a primary to coordinate write operation. • Two types of protocols: • Remote-Write Protocols (primary-backup) • Write operations submit to a single remote server. • Read op locally, • Local-Write Protocols (primary-backup w local writes) • the primary migrates to the process wanting to perform an update.

  8. Primary-based protocolsRemote-Write Protocols • Figure 7-20. The principle of a primary-backup protocol. • Long time to finish – blocking or nonblocking the initiator • Fault tolerance issue related • Primary does the ordering

  9. Primary-based protocolsLocal-Write Protocols • Multiple, successive writes can be loca • Nonblocking

  10. Consistency protocol • Primary-based protocols • Which protocol? • Example: Traditionally applied in distributed databases and file systems that require a high degree of fault tolerance. Replicas are often placed on same LAN. • Example: Mobile computing in disconnected mode (ship all relevant files to user before disconnecting, and update later on). Other processes can not update.

  11. Consistency protocol • Describes the implementation of a specific consistency model. Continuous consistence Sequential consistence Primary-based protocols Replicated-write protocols

  12. Replicated-write protocols • Write operations can be carried out at multiple replicas instead of one. • Active replication • Propagate the process (write operation) that cause the updates (in stead of the updates) to replica • Operations need to be carried in the same order everywhere. • Totally ordered Multicast or a central coordinator – sequencer. • Quorum-Based Protocols • Client need to request and acquire the permission of multiple servers before reading or writing a replicated data item

  13. Quorum-Based Protocols Quorum - the smallest number of people needed to be present at a meeting before it can officially begin and before official decisions can be taken. • Ensure that each operation is carried out in such a way that a majority vote is established • more than half of the N servers (and plus 1) • So to allow determine the consistency and perform the operation • Version numbers - Newer versions

  14. Quorum-Based Protocols • distinguish read quorum and write quorum • For update: When agreed, update and increase version of the data • For read: also need more than half of the N servers (and plus 1) to agree and to send data. and increase version of the data • How many data with same version? • But can be relaxed to • Nw>N/2 -- prevent write-write conflict • Nw + Nr > N -- prevent read-write conflict

  15. Replicated-write protocolsQuorum-Based Protocols • Three examples of the voting algorithm. (a) A correct choice of read and write set. (b) A choice that may lead to write-write conflicts. (c) A correct choice, known as ROWA (read one, write all).

More Related