OceanStore & GPFS - Revision history

Emmellst: /* GPFS */

2008-02-25T20:52:24Z

GPFS

← Older revision		Revision as of 20:52, 25 February 2008
Line 39:		Line 39:
	*Locking mechanism		*Locking mechanism
	**Two options		**Two options
	***Data shipping		***1. Data shipping
	****Distributed		****Distributed
	****First client to request access to file receives token		****First client to request access to file receives token
	****Other clients must request the current owner of the token		****Other clients must request the current owner of the token
	*****The current owner of the file grants portional access to their file (breaks token and gives portion access)		*****The current owner of the file grants portional access to their file (breaks token and gives portion access)
	***Centralized locking		***2. Centralized locking
	****Faster in a small disk circumstance		****Faster in a small disk circumstance
	*Extreme reliability		*Extreme reliability

Emmellst: /* GPFS */

2008-02-25T20:50:55Z

GPFS

← Older revision		Revision as of 20:50, 25 February 2008
Line 34:		Line 34:

	Pros		Pros
	**Massively parallel - data is striped across many many disks		*Massively parallel - data is striped across many many disks
	**Therefor read/write is very fast		*Therefor read/write is very fast
	*Option of redundancy		*Option of redundancy
	*Locking mechanism		*Locking mechanism
Line 53:		Line 53:
	*Everything must be trusted! Designed for clusters, not across LAN/WAN		*Everything must be trusted! Designed for clusters, not across LAN/WAN
	*Not appropriate for distributed networks.		*Not appropriate for distributed networks.


	=XUFS=		=XUFS=

Emmellst at 20:50, 25 February 2008

2008-02-25T20:50:32Z

← Older revision		Revision as of 20:50, 25 February 2008
Line 16:		Line 16:
	=Ocean Store=		=Ocean Store=
	Pros		Pros
	-Only trust required is own box		*Only trust required is own box
	-Data is highly durable due to file versioning		*Data is highly durable due to file versioning
	-Information divorced from location		*Information divorced from location
	--So long as you can reliably obtain information, it doesn't matter where it is located		**So long as you can reliably obtain information, it doesn't matter where it is located
	-Applicable to many data storage situations, not for a specific case		*Applicable to many data storage situations, not for a specific case
	-Routing is decentralized		*Routing is decentralized
	-2/3 of network is up? All is available		*2/3 of network is up? All is available

	Cons		Cons
	-Very expensive to computer cryptography (slow generation of keys)		*Very expensive to computer cryptography (slow generation of keys)
	-Utility models don't make economic sense, people prefer not to pay for access to their data		*Utility models don't make economic sense, people prefer not to pay for access to their data

	~~=Pond=~~
	~~Example of oceanstore~~


Line 35:		Line 32:
	Distributed local OS designed for clusters		Distributed local OS designed for clusters
	Max size of 4096TB		Max size of 4096TB

	Pros		Pros
	-Massively parallel - data is striped across many many disks		**Massively parallel - data is striped across many many disks
	--Therefor read/write is very fast		**Therefor read/write is very fast
	-Option of redundancy		*Option of redundancy
	-Locking mechanism		*Locking mechanism
	--Two options		**Two options
	~~---~~Data shipping		***Data shipping
	~~----~~Distributed		****Distributed
	~~----~~First client to request access to file receives token		****First client to request access to file receives token
	~~----~~Other clients must request the current owner of the token		****Other clients must request the current owner of the token
	~~-----~~The current owner of the file grants portional access to their file (breaks token and gives portion access)		*****The current owner of the file grants portional access to their file (breaks token and gives portion access)
	~~---~~Centralized locking		***Centralized locking
	~~----~~Faster in a small disk circumstance		****Faster in a small disk circumstance
	-Extreme reliability		*Extreme reliability
	--Able to literally remove a hotswap disk and insert a blank one in its place, only to have the blank disk completely regenerate the missing data		**Able to literally remove a hotswap disk and insert a blank one in its place, only to have the blank disk completely regenerate the missing data
	--Journalling to record token ownership - helps recovery when node in possession dies		**Journalling to record token ownership - helps recovery when node in possession dies

	Cons		Cons
	-Everything must be trusted! Designed for clusters, not across LAN/WAN		*Everything must be trusted! Designed for clusters, not across LAN/WAN
	-Not appropriate for distributed networks.		*Not appropriate for distributed networks.


	=XUFS=		=XUFS=
	User-space implementation		*User-space implementation
	Designed to be simple		*Designed to be simple
	Very generic		*Very generic

Emmellst at 20:46, 25 February 2008

2008-02-25T20:46:26Z

← Older revision		Revision as of 20:46, 25 February 2008
Line 58:		Line 58:

	=XUFS=		=XUFS=
	Designed ~~for OS with~~		User-space implementation
			Designed to be simple
			Very generic

Emmellst: /* Questions */

2008-02-25T20:43:42Z

Questions

← Older revision		Revision as of 20:43, 25 February 2008
Line 10:		Line 10:

	==Questions==		==Questions==
			Is it worth it??



			=Ocean Store=
			Pros
			-Only trust required is own box
			-Data is highly durable due to file versioning
			-Information divorced from location
			--So long as you can reliably obtain information, it doesn't matter where it is located
			-Applicable to many data storage situations, not for a specific case
			-Routing is decentralized
			-2/3 of network is up? All is available

			Cons
			-Very expensive to computer cryptography (slow generation of keys)
			-Utility models don't make economic sense, people prefer not to pay for access to their data

			=Pond=
			Example of oceanstore


			=GPFS=
			Distributed local OS designed for clusters
			Max size of 4096TB
			Pros
			-Massively parallel - data is striped across many many disks
			--Therefor read/write is very fast
			-Option of redundancy
			-Locking mechanism
			--Two options
			---Data shipping
			----Distributed
			----First client to request access to file receives token
			----Other clients must request the current owner of the token
			-----The current owner of the file grants portional access to their file (breaks token and gives portion access)
			---Centralized locking
			----Faster in a small disk circumstance
			-Extreme reliability
			--Able to literally remove a hotswap disk and insert a blank one in its place, only to have the blank disk completely regenerate the missing data
			--Journalling to record token ownership - helps recovery when node in possession dies

			Cons
			-Everything must be trusted! Designed for clusters, not across LAN/WAN
			-Not appropriate for distributed networks.


			=XUFS=
			Designed for OS with

Soma: /* Readings */

2008-02-22T20:08:47Z

Readings

← Older revision		Revision as of 20:08, 22 February 2008
Line 2:		Line 2:

	[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/oceanstore-sigplan.pdf John Kubiatowicz et al., "OceanStore: An Architecture for Global-Scale Persistent Storage" (2000)]		[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/oceanstore-sigplan.pdf John Kubiatowicz et al., "OceanStore: An Architecture for Global-Scale Persistent Storage" (2000)]

			[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/fast2003-pond.pdf Sean Rhea et al., "Pond: the OceanStore Prototype" (2003)]

	[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/gpfs-fast02.pdf Frank Schmuck and Roger Haskin, "GPFS: A Shared-Disk File System for Large Computing Clusters" (2002)]		[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/gpfs-fast02.pdf Frank Schmuck and Roger Haskin, "GPFS: A Shared-Disk File System for Large Computing Clusters" (2002)]

	~~[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/fast2003-pond.pdf Pond]~~		[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/walker-xufs-worlds06.pdf Edward Walker, "A Distributed File System for a Wide-Area High Performance Computing Infrastructure" (2006)]

	[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/walker-xufs-worlds06.pdf ~~XUFS~~]

	==Questions==		==Questions==

Soma: /* Readings */

2008-02-22T20:05:07Z

Readings

← Older revision		Revision as of 20:05, 22 February 2008
Line 3:		Line 3:
	[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/oceanstore-sigplan.pdf John Kubiatowicz et al., "OceanStore: An Architecture for Global-Scale Persistent Storage" (2000)]		[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/oceanstore-sigplan.pdf John Kubiatowicz et al., "OceanStore: An Architecture for Global-Scale Persistent Storage" (2000)]

	[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/gpfs-fast02.pdf GPFS]		[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/gpfs-fast02.pdf Frank Schmuck and Roger Haskin, "GPFS: A Shared-Disk File System for Large Computing Clusters" (2002)]

	[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/fast2003-pond.pdf Pond]		[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/fast2003-pond.pdf Pond]

Soma: Oceanstore & GPFS moved to OceanStore & GPFS

2008-02-22T19:59:10Z

Oceanstore & GPFS moved to OceanStore & GPFS

← Older revision	Revision as of 19:59, 22 February 2008
(No difference)

Soma: /* Readings */

2008-02-22T19:58:58Z

Readings

← Older revision		Revision as of 19:58, 22 February 2008
Line 1:		Line 1:
	==Readings==		==Readings==

	[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/oceanstore-sigplan.pdf ~~Oceanstore~~]		[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/oceanstore-sigplan.pdf John Kubiatowicz et al., "OceanStore: An Architecture for Global-Scale Persistent Storage" (2000)]

	[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/gpfs-fast02.pdf GPFS]		[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/gpfs-fast02.pdf GPFS]

Soma: /* Readings */

2008-02-22T19:57:32Z

Readings

← Older revision		Revision as of 19:57, 22 February 2008
Line 2:		Line 2:

	[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/oceanstore-sigplan.pdf Oceanstore]		[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/oceanstore-sigplan.pdf Oceanstore]

	[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/gpfs-fast02.pdf GPFS]		[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/gpfs-fast02.pdf GPFS]

	[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/fast2003-pond.pdf Pond]		[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/fast2003-pond.pdf Pond]

	[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/walker-xufs-worlds06.pdf XUFS]		[http://homeostasis.scs.carleton.ca/~soma/distos/2008-02-25/walker-xufs-worlds06.pdf XUFS]

	==Questions==		==Questions==