Amoeba A Distributed Operating System for the 1990s

1
Amoeba A Distributed Operating System for the
1990s

• Authors
• Sape J. Mullender,
• Guido van Rossum,
• Andrew S. Tanenbaum,
• Robbert van Renesse,
• and Hans van Staveren

Presented by Oliver Hampton
Wednesday, October 1, 2003
2
Amoeba

• Developed at
• Vrije Unviversiteit (Amsterdam)
• (Free University)
• In cooperation with
• Centrum voor Wiskunde en Informatica (Amsterdam)
• (Center for Mathematics and Computer Science)
• Research began in 1980

3
Components

• Workstations supports editing and other tasks
  that require fast interactive response
• Processor Pool does most of the heavy duty
  processing
• Specialized Servers file server, data base
• Gateway links Amoeba domains together

4
Network vs. Transparent Distributed System

• Network
• Each user logs into one specific machine (home
  machine).
• Programs execute on the home machine, unless the
  user gives an explicit command to run it
  elsewhere.
• Transparent Distributed System
• User logs into the system, not to any specific
  machine.
• When a program is run, the system, not the user
  determines the best place to run it.

5
Design Goals

• Transparency
• Distribution
• Parallelism
• High Performance

6
Objects and Capabilities

• Amoeba is an Object-Based system
• Object abstract data type on which well-defined
  operations may be performed
• Examples file, process, or directory
• Objects are identified and protected by a
  capability
• Capability a handle on an object. Allows the
  folder to perform operations on the object.

7
Communication Primitives

• Remote Procedure Call (RPC) model is used between
  Client and Server
• do_operation
• get_request
• send_reply
• Multiple Inheritance via Classes
• Amoeba Interface Language (AIL)
• Interface for object manipulation
• Marshal / Unmarshal RPC parameters
• Request / Reply transport mechanism

8
Now that we know what objects are, how do we find
them?

• Objects in Amoeba may be physically located
  anywhere there is disk space.
• Locating
• do_operation call arrives at kernel
• Kernel checks if Service Port is known
• If not, kernel broadcasts locate packet
• Locate packet asks if there are servers with
  get_requests for the port in question

9
Amoeba File System

• Public capabilities are accessible by users
• Such as command executables, public files, data
  bases
• Hierarchical Directory Structure
• Directories are Objects
• Directory is a set of (name, capability) pairs
• Basic Operations
• Lookup
• Enter
• Delete
• Directory Server
• Reliability Dependability
• Security directories may be encrypted

10
Bullet File Server

• Increased Performance - FAST
• 3 fold increase over Sun Network File System
  (NFS)
• Files are stored contiguously (next slide)
• Increased Fragmentation
• Example may require 800 MB disk to store 500 MB
  worth of files
• Files are immutable
• Read_file
• Create_file
• Delete_file

11
Bullet File Server Continued

• Stores files contiguously on disk and in servers
  RAM cache
• Processors may only operate on files that fit in
  their physical memory

12
Process Management

• A Process is an Object
• Process information is contained in Capabilities
  and in a data structure called the Process
  Descriptor

• Process Descriptor
• Host Descriptor
• Capabilities
• Segment Component
• Thread Component

13
Process Management Continued

• Processes have explicit control over their
  address space.
• mapping segments in and out
• Processes States
• Running
• Stunned
• Process exists but does not execute instructions
  example, debugging

14
Process Management Continued

• Threads
• Process start-up it has at least one thread
• Thread number is dynamic
• During execution
• Process may create additional threads
• Process may terminate existing threads
• Threads are managed by the kernel
• When threads do RPCs
• Kernel can block thread
• Kernel can schedule another thread in blocked
  threads place if one is available

15
Performance

• Amoeba performance data on two 16.7-megahertz
  Motorola MC68020s, with a user process on each,
  communicating over a 10-megabit-per-second
  Ethernet.
• Native Amoeba Remote Communication at 8 KB (Case
  2) is 3.05 times faster than the Sun RPC remote
  (NFS) and 2.78 times faster than the Unix driver
  remote.

16
Conclusions

• Review of Design Goals
• Transparency
• User is unaware of location and number of
  processors available in Amoeba System
• Distribution
• Objects are distributed, as is computational
  power
• Parallelism
• Multi-Thread Processes
• High Performance
• Bullet File Server, performance data