Leveraging Operational Transformation in Collaborative Object Groups

1
Leveraging Operational Transformationin
Collaborative Object Groups

• G. Henri ter Hofteterhofte_at_telin.nl
• Telematica Instituut, The Netherlands

2
It takes 100 ms ...

• for humans to react, or for humans to perceive a
  response is not immediate
• for 40 million clock cycles to run on a 400 MHz
  Pentium II
• for 15,5 Mb of data to be sent on an ATM link
• for 1 bit of that data to reach its destination
  on an international link
• (for light to travel about 3/4 around the world)
• latency it wont go away!

3
What if

• Latency gt 100 ms and you need immediate
  response?
• replication!
• for performance of groupware (despite high
  latency)
• You want to work together while disconnected?
• replication!
• for availability in nomadic computing (i.e.
  despite disconnection)
• A computer crashes and you need the
  functionality?
• replication!
• for fault-tolerance in critical applications
  (i.e. despite partial faults)

4
Collaborative zipper
file status
Legend
edit status
uncoupled status
view control status
coupled status
(centralized)
user interface control status
coupled status
(synchronization)
5
Replication complexities

• consistency management
• due to ordering anomalies
• waiting on the other side(s) is expensive
• e.g. dont lock unless really necessary!
• latecomers earlyleavers
• how to handle without disrupting service?

6
Remote Method Invocation
B
7
Multicast RMI ?
JVM3
JVM2
JVM1
B2
A
B1
Multicast RMI (MRMI)
8
Object group invocation
9
Object group invocation
10
Proposed solution GroupRMI

• replicated object groups
• programmers can treat them as normal
  (non-replicated) objects
• replication transparency
• adhere to Java RMI, works for arbitrary remote
  objects
• programmers that know how to RMI, know how to
  GroupRMI
• shallow learning curve
• default policies allow for policy plug-ins
• to allow advanced programmers to deal with
  special cases
• to allow various (optimistic) consistency
  policies
• coupling/uncoupling, operational transformation,
  merging
• selective replication transparency

11
One level
Legend
uncoupled status
coupled status
12
One level
Legend
uncoupled status
coupled status
13
One object group
Legend
uncoupled status
coupled status
14
One object group
Legend
uncoupled status
coupled status
15
One object group member
16
One object group member
17
Java RMI scenario
client JVM
server JVM (h1)
Registry
Naming
Naming
A
18
GroupRMI scenario
JVM
JVM
19
Operational transformation matrix two operations

• commutative do nothing special
• masking delete one from history buffer
• memoryless masking all, empty buffer
• conflicting specify transformation function
• application specific
• fork history (parallel versions) is an
  optionfollowed by version merging tools

20
Use of operational transformation

• State transfer intentional latency
• Group undo/redo delayed undo
• Version merging human-assisted form of
  operational transformation

21
Issues

• terminology
• find better quality criteria
• e.g. from serializability to consistency
  causality, intention preservation
• performance studies
• e.g. impact of latency, use of IP multicast,
  algorithms
• simulation of network
• usability tests
• what DO users expect, what IS their intention
• build application-specify transformation matrices

22
Issues (contd)

• share algorithms
• common experimentation platform
• with pluggable policies
• simulation

23
Thats all folks!

• Questions? Comments?