Things you should know about operating systems

1
Things you should know about operating systems

• Yasushi Saito

2
Executive Summary

• Whats dead
• Microkernel
• File systems
• Distributed OS
• Persistent OS
• Single address space OS
• Distributed checkpointing/rollbacking
• Distributed shared memory
• Extensible OS

• Whats hot
• Web-related tuning
• Distributed caching
• Multimedia
• Clustering
• Storage service
• Binary rewriting, machine emulation
• Mobility adaptation

Trend towards solving real problems
3
Web stuff

• OS behavior under very heavy network load
• Zero-copy I/O fbufs, I/O lite (Pai/Drushchel_at_rice
  ), sendfile(2) (many commercial UNIX)
• Decouple process mgmt resource mgmt (really not
  specific to web)
• Eg, fair-share-schedule server and CGI as a unit
• Performance isolation (Verghese_at_WRL), resource
  container (Banga/Druschel_at_rice), Rialto
  (Jones_at_msr), etc

4
Web stuff

• Polling vs interrupt
• interrupt can thrash the server under heavy load
• Mogul/Ramakrishnan _at_WRL, Soft timers
  (Aron/Druschel_at_Rice)
• Hacking select(2), accept(2) (Banga/Druchel)
• Squid spends 50 of CPU there
• Clustering, load balancing Fox_at_UCB/Stanford,
  LARD(Rice)

5
Distributed Caching

• Why?
• CPU network are fast, but disk is slow
• GMS (Feeley/Voelker_at_UW), xFS(tom_at_Berkeley)
  Cooperative file caching
• Petal/Frangipani(SRC) Distributed disk
• Cooperative web caching (Wolman/Voelker_at_UW)
  limited benefit (log hit-rate increase wrt client
  population)

6
Clustering

• Two uses number crunching and Internet serving
• Number crunching
• NOW (Berkeley)
• Internet service
• Driven by commercial demand, research issues not
  really understood
• Database cluster (many)
• Web server cluster (many)
• Porcupine(me) automatic failover by dynamic data
  distribution

7
Devices

• Add brain to device
• Active disks(many) sorting, joining, etc
• SPINE(mef_at_UW) programmable network
• IRAM (Berkeley) compute on memory
• Misc
• Petal (SRC) distributed virtual disk w/ failover
  and load balancing
• Virtual LFS disk (Wang_at_Berkeley/Princeton)
• Logical volume management, system area
  network(SAN)

8
Binary rewriting, machine emulation, etc

• Virtual machine emulation (Rosenblum_at_Stanfond)
• SimOS, VMware emulates the entire machine
• Disco run SMP OS on NUMA cluster
• Binary rewriting
• Eraser (Savage_at_UW) dynamic race detection
• Etch (UW) atom for i386
• Kimera (egs_at_UW) distributed Java virtual machine
• Shasta (SRC) DSM converter for commercial SMP
  app
• Low-level performance measurement
• DCPI (SRC/WRL) pinpoint cache miss, pipeline
  stall, etc

9
Mobility/Adaptation

• Adapting to changing network quality, battery
  life Odyssey (Satya_at_CMU)
• changing image resolution, sound quality,
  controlling screen lighting
• Mobile database
• low-bandwidth communication
• any-to-any reconciliation
• Bayou (PARC), Roam (UCLA)

10
Multimedia support

• CPU scheduling
• Combining real time and fair-share scheduling
• BVT, SMART (Duda,Nieh_at_Stanford)
• File system
• extent-based allocation
• rate guarantee
• Clustering to improve load balance
• Tiger (Bolosky_at_MSR), RIO (UCLA)

11
Local interest

• Alec Wolman cooperative web caching
• E. Gun Sirer Kimera
• Marc Fuiczynski SPINE
• Ashutosh Tiwary Perf. evaluation of OODB
• Przemek Pardyak Benchmark generation
• Yasushi Saito Porcupine

12
Whats dead?

• Micro/distributed/persistent/extensible OS
• No killer apps
• File systems
• People are satisfied
• DSMcheckpointing
• Degenerated into minor protocol tweaking

• Exceptions
• Elephant (UBC) file system that never forgets
• Shasta (SRC) converts MT app to DSM app.

13
Conference/Workshop to Watch For

• Symp. on Op. Sys. Principles (SOSP)
• Op. Sys. Design and Impl. (OSDI)
• USENIX Technical Conf. (USENIX)
• USENIX Symp. on Internet Tech. and Sys. (USITS)
• ASPLOS
• ISCA, SC, SIGMOD, VLDB, SIGCOMM, FTCS, PLDI, .
• Workshop on Hot topics in Op. Sys SIGOPS
  European workshop