L09: Reliability in DDBMS

1
L09 Reliability in DDBMS
2
Reliability of DDBMS

• Atomicity and durability of transactions
• Ability to continue processing user requests when
  the underlying system is unreliable
• Dependent on the reliability of the hardware and
  the software components that make up the
  distributed environment

3
Formal Definitions

• The reliability of a system, R(t) at time t, is
  defined as a probability function
• R(t) Pr0 failures in 0,t 0 failures at t
  0
• The availability of a system, A(t) at time t, is
  the probability that the system is operational
  according to its specification at time t.

4
System and State

• System
• A collection of components interacting w/ env.
• Design how components are put together
• States
• External state response a system gives to env.
• Internal state union of external states among
  components
• Specification valid behavior of each state

5
Failures, Errors, and Faults

• Failures
• Any deviation of a system from the behavior
  described in the specification
• Erroneous states (error the incorrect part)
• Internal states that do not obey the
  specification
• Fault
• Any error in the internal states or in the design

6
Failures in Distributed Systems

• SLAC Data
• HW 57, SW 12, Operations 14, and Environment
  17
• Tandem Data
• HW 18, SW 25, Maintenance 25, Operations 17,
  and Environment 15
• 5ESS Switch Data
• HW 32.3, SW 44.3, Operations 17.5, etc.

7
MTBF and MTTR

• Mean Time Between Failures (MTBF)
• Expected time between subsequent failures
• MTBF ? R(t)dt
• Mean Time To Repair (MTTR)
• Expected time to repair a failed system
• Availability MTBF/ (MTBF MTTR)

8
Occurrence of Events over Time
MTBF
MTTR
MTTD
Time
Fault occurs
Error caused
Detection of error
Fault occurs
Error caused
Repair
Mean Time To Detect
9
Failures in Distributed DBMS

• Transaction failures (aborts)
• Error in inputs, deadlocks resolution, CC aborts
• Site (System) failures
• Hardware (power, memory, CPU) or software
• Media (disk) failures
• Communication failures
• Errors in messages, improperly ordered msgs, lost
  or undeliverable msgs, and line failures

10
Distributed Reliability Protocols

• Commit protocols
• How to execute commit command for distributed
  transactions
• Termination protocols
• If one site fails, how do the other sites
  terminate the transaction?
• Recovery protocols
• How to execute the recover command for
  distributed transactions

11
TwoPhase Commit (2PC)

• Phase 1
• The coordinator sends prepare messages to the
  participants and the participants reply yes or
  no
• Phase 2
• The coordinator decides to commit or abort the
  transaction and send the decision to all
  participants
• The participants execute the decision and send
  back acknowledgement

12
Global Commit Rule of 2PC

• If at least one participant vote to abort the
  transaction, the coordinator has to reach a
  global abort decision.
• If all of the participants vote to commit the
  transaction, the coordinator has to reach a
  global commit decision.

13
Coordinator
Participant
INITIAL
INITIAL
PREPARE
Write begin-commit in log
Ready to commit?
Wirte abort in log
No
Vote ABORT
WAIT
Vote COMMIT
Yes
ABORT
Yes
Any No?
Wirte ready in log
Wirte abort in log
GLOBAL-ABORT
No
READY
Wirte commit in log
GLOBAL-COMMIT
ABORT
Type of msg?
Abort
ACK
COMMIT
Wirte abort in log
Commit
ACK
Wirte cmmit in log
Wirte EOT in log
ABORT
COMMIT
2PC Protocol Actions
14
Communications in 2PC

• Centralized 2PC (as depicted in previous slides)
• Communications are between the coordinator and
  the participants
• No communications are between the participants
• Linear 2PC (or called nested 2PC)
• Participants communicate with one another in
  order
• Distributed 2PC
• Each participant broadcasts its vote to all
  participants
• No need for the second phase

15
Variants of 2PC

• Goal performance
• Number of msgs and number of log writes
• Variations
• Presumed abort 2PC whenever there is no
  information about the xact, it was an abort.
• Presumed commit 2PC whenever there is no
  information about the xact, it was a commit.

16
State Transitions in 2PC
INITIAL
Participants
Prepare Vote-commit
Prepare Vote-abort
Coordinator
INITIAL
READY
Global-abort Ack
Commit command Prepare
Global-commit Ack
WAIT
COMMIT
ABORT
Vote Commit (all) Global-commit
Vote abort (any) Global-abort
COMMIT
ABORT
17
Failures at a Participating Site

• The coordinator C detects that a site is failed
• before a vote message is received assume
  vote-abort
• after a vote message is received ignore the
  failure

• A participating site recovers from a failure
• The log contains commit/abort record the site
  execute redo/undo
• The log contains ready record query the status
  of the transaction, redo or undo accordingly
• The log contains no such control record undo

18
Failures at Coordinator Site

• If the coordinator fails, the participants must
  decide the fate of the transaction T
• If an active site contains commit/abort in its
  log, T must be committed/aborted
• If some active site does not contain ready record
  in its log, it is preferable to abort T
• If the status of T cannot be determined, it has
  to wait for the coordinator to recover.

19
Recovery Protocols -- Motivation

• Atomicity
• Transactions may abort (Rollback).
• Durability
• What if DBMS stops running? (Causes?)

• Desired Behavior after system restarts
• T1, T2 T3 should be durable.
• T4 T5 should be aborted (effects not seen).

crash!
T1 T2 T3 T4 T5
20
Assumptions

• Concurrency control is in effect.
• Strict 2PL, in particular.
• Updates are happening in place.
• i.e. data is overwritten on (deleted from) the
  disk.
• A simple scheme to guarantee Atomicity
  Durability?

21
Basic Idea Logging

• Record REDO and UNDO information, for every
  update, in a log.
• Sequential writes to log (put it on a separate
  disk).
• Minimal info (difference) written to log, so
  multiple updates fit in a single log page.
• Log An ordered list of REDO/UNDO actions
• Log record contains
• ltXID, pageID, offset, length, old data, new datagt
  
• and additional control info.

22
Write-Ahead Logging (WAL)

• The Write-Ahead Logging Protocol
• Must force the log record for an update before
  the corresponding data page gets to disk.
• Must write all log records for a transaction
  before commit.
• 1 guarantees Atomicity.
• 2 guarantees Durability.

23
WAL the Log

• Each log record has a unique Log Sequence Number
  (LSN).
• LSNs always increasing.
• Each data page contains a pageLSN.
• The LSN of the most recent log record
  for an update to
  that page.
• System keeps track of flushedLSN.
• The max LSN flushed so far.
• WAL Before a page is written,
• pageLSN flushedLSN

Log records flushed to disk
Log tail in RAM
24
Log Records

• Possible log record types
• Update
• Commit
• Abort
• End (signifies end of commit or abort)
• Compensation Log Records (CLRs)
• for UNDO actions

LogRecord fields
update records only
25
Other Log-Related State

• Transaction Table
• One entry per active transacion.
• Contains XID, status (running/committed/aborted),
  and lastLSN.
• Dirty Page Table
• One entry per dirty page in buffer pool.
• Contains recLSN -- the LSN of the log record
  which first caused the page to be dirty.

26
Normal Execution of a Transaction

• Series of reads writes, followed by commit or
  abort.
• We will assume that write is atomic on disk.
• In practice, additional details to deal with
  non-atomic writes.
• Strict 2PL.
• STEAL, NO-FORCE buffer management, with
  Write-Ahead Logging.

27
Checkpointing

• Periodically, the DBMS creates a checkpoint, in
  order to minimize the time taken to recover in
  the event of a system crash. Write to log
• begin_checkpoint record Indicates when chkpt
  began.
• end_checkpoint record Contains current Xact
  table and dirty page table. This is a fuzzy
  checkpoint
• Other Xacts continue to run so these tables
  accurate only as of the time of the
  begin_checkpoint record.
• No attempt to force dirty pages to disk
  effectiveness of checkpoint limited by oldest
  unwritten change to a dirty page. (So its a good
  idea to periodically flush dirty pages to disk!)
• Store LSN of chkpt record in a safe place (master
  record).

28
The Big Picture Whats Stored Where
LOG
RAM
DB
LogRecords
Xact Table lastLSN status Dirty Page
Table recLSN flushedLSN
Data pages each with a pageLSN
master record
29
Simple Transaction Abort

• For now, consider an explicit abort of a Xact.
• No crash involved.
• We want to play back the log in reverse order,
  UNDOing updates.
• Get lastLSN of Xact from Xact table.
• Can follow chain of log records backward via the
  prevLSN field.
• Before starting UNDO, write an Abort log record.
• For recovering from crash during UNDO!

30
Abort, cont.

• To perform UNDO, must have a lock on data!
• No problem!
• Before restoring old value of a page, write a
  CLR
• You continue logging while you UNDO!!
• CLR has one extra field undonextLSN
• Points to the next LSN to undo (i.e. the prevLSN
  of the record were currently undoing).
• CLRs never Undone (but they might be Redone when
  repeating history guarantees Atomicity!)
• At end of UNDO, write an end log record.

31
Transaction Commit

• Write commit record to log.
• All log records up to Xacts lastLSN are flushed.
• Guarantees that flushedLSN ³ lastLSN.
• Note that log flushes are sequential, synchronous
  writes to disk.
• Many log records per log page.
• Commit() returns.
• Write end record to log.

32
Crash Recovery Big Picture
Oldest log rec. of Xact active at crash

• Start from a checkpoint (found via master
  record).
• Three phases. Need to
• Figure out which Xacts committed since
  checkpoint, which failed (Analysis).
• REDO all actions.
• (repeat history)
• UNDO effects of failed Xacts.

Smallest recLSN in dirty page table after Analysis
Last chkpt
CRASH
A
R
U
33
Recovery The Analysis Phase

• Reconstruct state at checkpoint.
• via end_checkpoint record.
• Scan log forward from checkpoint.
• End record Remove Xact from Xact table.
• Other records Add Xact to Xact table, set
  lastLSNLSN, change Xact status on commit.
• Update record If P not in Dirty Page Table,
• Add P to D.P.T., set its recLSNLSN.

34
Recovery The REDO Phase

• We repeat History to reconstruct state at crash
• Reapply all updates (even of aborted Xacts!),
  redo CLRs.
• Scan forward from log rec containing smallest
  recLSN in D.P.T. For each CLR or update log rec
  LSN, REDO the action unless
• Affected page is not in the Dirty Page Table, or
• Affected page is in D.P.T., but has recLSN gt LSN,
  or
• pageLSN (in DB) ³ LSN.
• To REDO an action
• Reapply logged action.
• Set pageLSN to LSN. No additional logging!

35
Recovery The UNDO Phase

• ToUndo l l ? lastLSN of a loser Xact
• Repeat
• Choose largest LSN among ToUndo.
• If this LSN is a CLR and undonextLSNNULL
• Write an End record for this Xact.
• If this LSN is a CLR, and undonextLSN ! NULL
• Add undonextLSN to ToUndo
• Else this LSN is an update. Undo the update,
  write a CLR, add prevLSN to ToUndo.
• Until ToUndo is empty.

36
Example of Recovery
LSN LOG
begin_checkpoint end_checkpoint update T1
writes P5 update T2 writes P3 T1 abort CLR Undo
T1 LSN 10 T1 End update T3 writes P1 update T2
writes P5 CRASH, RESTART
00 05 10 20 30 40
45 50 60
prevLSNs
Xact Table lastLSN status Dirty Page
Table recLSN flushedLSN
ToUndo
37
Example Crash During Restart!
LSN LOG
00,05 10 20 30 40,45 50
60 70 80,85 90
begin_checkpoint, end_checkpoint update T1
writes P5 update T2 writes P3 T1 abort CLR Undo
T1 LSN 10, T1 End update T3 writes P1 update T2
writes P5 CRASH, RESTART CLR Undo T2 LSN 60 CLR
Undo T3 LSN 50, T3 end CRASH, RESTART CLR Undo
T2 LSN 20, T2 end
undonextLSN
Xact Table lastLSN status Dirty Page
Table recLSN flushedLSN
ToUndo
38
Additional Crash Issues

• What happens if system crashes during Analysis?
  During REDO?
• How do you limit the amount of work in REDO?
• Flush asynchronously in the background.
• Watch hot spots!
• How do you limit the amount of work in UNDO?
• Avoid long-running Xacts.

39
Summary of Logging/Recovery

• Recovery Manager guarantees Atomicity
  Durability.
• Use WAL to allow STEAL/NO-FORCE w/o sacrificing
  correctness.
• LSNs identify log records linked into backwards
  chains per transaction (via prevLSN).
• pageLSN allows comparison of data page and log
  records.
• Checkpointing A quick way to limit the amount
  of log to scan on recovery.
• Recovery works in 3 phases
• Analysis Forward from checkpoint.
• Redo Forward from oldest recLSN.
• Redo repeats history Simplifies the logic!
• Undo Backward from end to first LSN of oldest
  Xact alive at crash.
• Upon Undo, write CLRs.

40
Summary, Cont.

• Checkpointing A quick way to limit the amount
  of log to scan on recovery.
• Recovery works in 3 phases
• Analysis Forward from checkpoint.
• Redo Forward from oldest recLSN.
• Undo Backward from end to first LSN of oldest
  Xact alive at crash.
• Upon Undo, write CLRs.
• Redo repeats history Simplifies the logic!