Storage Systems Dependability
Reza Eftekhary
eftekhary@ce.sharif.edu
In
the name of GOD
Storage Architecture
2
Storage Systems Dependability
Dependability Criteria
D
a
t
a
r
e
l
i
a
b
i
l
i
t
y
F
a
u
l
t
t
o
l
e
r
a
n
t
t
e
c
h
n
i
q
u
e
s
E
r
r
o
r
d
e
t
e
c
t
i
o
n
&
C
o
r
r
e
c
t
i
o
n
M
o
d
u
l
a
r
R
e
d
u
n
d
a
n
c
y
E
r
r
o
r
R
e
c
o
v
e
r
y
C
h
e
c
k
p
o
i
n
t
s
:
b
a
c
k
u
p
l
e
v
e
l
R
e
s
t
a
r
t
s
o
f
t
w
a
r
e
,
r
e
s
t
o
r
e
d
a
t
a
:
t
a
s
k
l
e
v
e
l
D
a
t
a
a
v
a
i
l
a
b
i
l
i
t
y
3
Storage Systems Dependability
Level of Evaluation
B
a
c
k
-
u
p
l
e
v
e
l
C
o
m
p
o
n
e
n
t
l
e
v
e
l
D
i
s
k
P
r
o
c
e
s
s
o
r
s
E
n
c
l
o
s
u
r
e
I
n
t
e
r
c
o
n
n
e
c
t
s
,
…
4
Storage Systems Dependability
Component Level-Disk
Type of faults
Protection techniques
Reliability Model
5
Storage Systems Dependability
Disk
Types of fault
Operational Failures (Can not find data)
Bad servo-track
Bad electronics
Can't stay on track
Bad read head
6
Storage Systems Dependability
Disk
Types of fault
Latent Defects (Data missing)
Error during writing
Bad media
Inherent bit-error rate
High-fly writes
7
Storage Systems Dependability
Disk
Type of fault
UnDetected Error (UDE)
Undetected Write Error
Undetected Read Error
Low Rate of occurrence
High Rate in Large storage system
UWE
Departed write
Near off-track write
Far off-track write
8
Storage Systems Dependability
Protection Techniques
ECC
Parity
Raid
Disk scrubbing
Intra-disk redundancy
9
Storage Systems Dependability
Redundant Array of Independent Disks
Data striping
10
Storage Systems Dependability
Double Disk Failure
Two scenarios result in DDF
two simultaneous operational failures
an operational failure occurs after a latent defect
before latent error corrected
Multiple simultaneous latent defects don’t constitute
failure.
11
Storage Systems Dependability
Disk Scrubbing
A background process
Reading sectors data and recover data via
ECC
Parity of other disks in RAID
Benefits
Decrease DDF
Increase reliability
Problems
Decrease bandwidth
Increase access to data
12
Storage Systems Dependability
Intra-Disk Redundancy
Increase redundancy information
2 or 3 fault tolerant codes
Store parity in row & column
Storing overheads
Decrease BW overhead
Less delay in comparison with scrubbing
13
Storage Systems Dependability
Backup Level
14
Storage Systems Dependability
Backup Level
Sample failure types
Storage hardware failures
disk drive, disk array, file server, rack, building, site, region
Common shared infrastructure failures
power, air condition
Software-induced data corruption
storage device firmware, application defects, virus
Human/operator/user error
accidental deletion or overwriting, external attacks
15
Storage Systems Dependability
Backup Level
Data protection techniques
Inside one storage device
partial redundancy or full redundancy
Between storage devices
mirroring at the same level
array to tape, or array to slow storage, across different levels
Update frequency:
synchronous, lock-step with foreground updates
continuous asynchronous - remote mirroring, logging
batched (e.g., nightly backup)
16
Storage Systems Dependability
Backup Level
Different tradeoffs for protection techniques
Protection overhead
including capacity, performance impact
on the foreground workload,
and downtime
Recovery time
how long does it take to get the data back?
Recovery points
how much (and which) data can be returned
to its pre-loss state?
Cost
Direct costs include equipment purchase or rental costs
Indirect costs- lost worker productivity, user dissatisfaction
17
Storage Systems Dependability
Backup Level
Requirement knowledge
The required bandwidth to mirror data
The number of the network links between the two sites
How often to take full backups and make incremental
The bandwidth required to make backups
When to move backup tapes off-site for disaster recovery
Whether to rebuild a stricken site, or build a new one
18
Storage Systems Dependability
Backup Level
When a failure occurs, it may be necessary to revert
back to a consistent point prior to the failure, which
will entail the loss of any data written after that point.
19
Storage Systems Dependability
Backup Level
Recovery time
elapsed time after a failure, that system
is
running again
the
recovery time objective (RTO)provides an acceptable
upper
bound
Recent data loss
amount of recent updates lost during recovery from a
failure
the recovery point objective (RPO) provides an upper
bound
20
Storage Systems Dependability
Backup Level
hierarchy in Primary and secondary copies
Primary data copy is level 0
Each level is responsible for
retaining some number of
discrete retrieval points
propagating RPs to the
next level in the hierarchy
21
Storage Systems Dependability
Backup Level
As the level increase, the data protection techniques
Store less frequent RPs
Present larger retention capacity
Exhibit longer recovery latencies
Recovery path
Opposite of RP propagation
22
Storage Systems Dependability
Backup Level
23
Storage Systems Dependability
Reference
[1]Jon G. Elerath. A Simple Equation for Estimating Reliability of an N+1
Redundant Array of Independent Disks (RAID). In Proceedings of the
International Conference on Dependable Systems and Networks (DSN), pages
484-493,2009.
[2]Jon G. Elerath and Michael Pecht. Enhanced reliability modeling of raid
storage systems. In Proceedings of the International Conference on
Dependable Systems and Networks (DSN), pages 175-184,2007.
[3]E.W.D.Rozier1, W.Belluomini1, V.Deenadhayalan1, J.Hafner1, KK.Rao, P.Zhou1.
Evaluating the Impact of Undetected Disk Errors in RAID Systems. In
Proceedings of the International Conference on Dependable Systems and
Networks (DSN), pages 83-92,2009.
[4]K. Keeton and A. Merchant. A framework for evaluating storage System
dependability. In Proc. Intl. Conf. on Dependable Systems and Networks
(DSN), pages 877–886, 2004.
[5] S. Seshadri , L. Chiu, C. Constantinescu, S. Balachandran, C. Dickey, L. Liu,
and P. Muench , ``Enhancing storage system availability on multi-core
architectures using recovery conscious scheduling,'' in USENIX FAST, 2008.
24
This content discusses storage systems dependability, covering topics such as data reliability, fault-tolerant techniques, error detection and correction, component levels, disk protection techniques, types of disk faults, and protection mechanisms like RAID and ECC. It provides insights into ensuring the reliable operation of storage systems.
Download Presentation
Please find below an Image/Link to download the presentation.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.
You are allowed to download the files provided on this website for personal or commercial use, subject to the condition that they are used lawfully. All files are the property of their respective owners.
The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author.
E N D
Presentation Transcript
In the name of GOD Reza Eftekhary eftekhary@ce.sharif.edu
Storage Architecture Storage Systems Dependability 2
Dependability Criteria Data reliability Fault tolerant techniques Error detection & Correction Modular Redundancy Error Recovery Checkpoints : backup level Restart software, restore data: task level Data availability Storage Systems Dependability 3
Level of Evaluation Back-up level Component level Disk Processors Enclosure Interconnects, Storage Systems Dependability 4
Component Level-Disk Type of faults Protection techniques Reliability Model Storage Systems Dependability 5
Disk Types of fault Operational Failures (Can not find data) Bad servo-track Bad electronics Can't stay on track Bad read head Storage Systems Dependability 6
Disk Types of fault Latent Defects (Data missing) Error during writing Bad media Inherent bit-error rate High-fly writes Storage Systems Dependability 7
Disk Type of fault UnDetected Error (UDE) Undetected Write Error Undetected Read Error Low Rate of occurrence High Rate in Large storage system UWE Departed write Near off-track write Far off-track write Storage Systems Dependability 8
Protection Techniques ECC Parity Raid Disk scrubbing Intra-disk redundancy Storage Systems Dependability 9
Redundant Array of Independent Disks Data striping Storage Systems Dependability 10
Double Disk Failure Two scenarios result in DDF two simultaneous operational failures an operational failure occurs after a latent defect before latent error corrected Multiple simultaneous latent defects don t constitute failure. Storage Systems Dependability 11
Disk Scrubbing A background process Reading sectors data and recover data via ECC Parity of other disks in RAID Benefits Decrease DDF Increase reliability Problems Decrease bandwidth Increase access to data Storage Systems Dependability 12
Intra-Disk Redundancy Increase redundancy information 2 or 3 fault tolerant codes Store parity in row & column Storing overheads Decrease BW overhead Less delay in comparison with scrubbing Storage Systems Dependability 13
Backup Level Storage Systems Dependability 14
Backup Level Sample failure types Storage hardware failures disk drive, disk array, file server, rack, building, site, region Common shared infrastructure failures power, air condition Software-induced data corruption storage device firmware, application defects, virus Human/operator/user error accidental deletion or overwriting, external attacks Storage Systems Dependability 15
Backup Level Data protection techniques Inside one storage device partial redundancy or full redundancy Between storage devices mirroring at the same level array to tape, or array to slow storage, across different levels Update frequency: synchronous, lock-step with foreground updates continuous asynchronous - remote mirroring, logging batched (e.g., nightly backup) Storage Systems Dependability 16
Backup Level Different tradeoffs for protection techniques Protection overhead including capacity, performance impact on the foreground workload, and downtime Recovery time how long does it take to get the data back? Recovery points how much (and which) data can be returned to its pre-loss state? Cost Direct costs include equipment purchase or rental costs Indirect costs- lost worker productivity, user dissatisfaction Storage Systems Dependability 17
Backup Level Requirement knowledge The required bandwidth to mirror data The number of the network links between the two sites How often to take full backups and make incremental The bandwidth required to make backups When to move backup tapes off-site for disaster recovery Whether to rebuild a stricken site, or build a new one Storage Systems Dependability 18
Backup Level When a failure occurs, it may be necessary to revert back to a consistent point prior to the failure, which will entail the loss of any data written after that point. Storage Systems Dependability 19
Backup Level Recovery time elapsed time after a failure, that system is running again the recovery time objective (RTO)provides an acceptable upper bound Recent data loss amount of recent updates lost during recovery from a failure the recovery point objective (RPO) provides an upper bound Storage Systems Dependability 20
Backup Level hierarchy in Primary and secondary copies Primary data copy is level 0 Each level is responsible for retaining some number of discrete retrieval points propagating RPs to the next level in the hierarchy Storage Systems Dependability 21
Backup Level As the level increase, the data protection techniques Store less frequent RPs Present larger retention capacity Exhibit longer recovery latencies Recovery path Opposite of RP propagation Storage Systems Dependability 22
Backup Level Storage Systems Dependability 23
Reference [1]Jon G. Elerath. A Simple Equation for Estimating Reliability of an N+1 Redundant Array of Independent Disks (RAID). In Proceedings of the International Conference on Dependable Systems and Networks (DSN), pages 484-493,2009. [2]Jon G. Elerath and Michael Pecht. Enhanced reliability modeling of raid storage systems. In Proceedings of the International Conference on Dependable Systems and Networks (DSN), pages 175-184,2007. [3]E.W.D.Rozier1, W.Belluomini1, V.Deenadhayalan1, J.Hafner1, KK.Rao, P.Zhou1. Evaluating the Impact of Undetected Disk Errors in RAID Systems. In Proceedings of the International Conference on Dependable Systems and Networks (DSN), pages 83-92,2009. [4]K. Keeton and A. Merchant. A framework for evaluating storage System dependability. In Proc. Intl. Conf. on Dependable Systems and Networks (DSN), pages 877 886, 2004. [5] S. Seshadri , L. Chiu, C. Constantinescu, S. Balachandran, C. Dickey, L. Liu, and P. Muench , ``Enhancing storage system availability on multi-core architectures using recovery conscious scheduling,'' in USENIX FAST, 2008. 24
