SWENG 541: Advanced Database Design Concepts

Crash Recovery
Log Checkpoint Recovery

Engineering Divsion, Penn State University Last Updated: November 17, 2004.

SWENG 541: Advanced Database Design Concepts

The ACID Properties

ENGINEERING DIVISION, PENN STATE UNIVERSITY

Atomicity: All actions in the Xact happen, or none happen. Consistency: If each Xact is consistent, and the DB starts

consistent, it ends up consistent.

Isolation:
Xacts.

Execution of one Xact is isolated from that of other If a Xact commits, its effects persist.

Durability:

The Recovery Manager guarantees Atomicity & Durability.

2

SWENG 541: Advanced Database Design Concepts

Motivation

ENGINEERING DIVISION, PENN STATE UNIVERSITY

Atomicity:
Transactions

may abort (Rollback).

Durability:
What

if DBMS stops running? (Causes?)

T1 T2 T3 T4 T5

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

crash!

SWENG 541: Advanced Database Design Concepts

Assumptions

ENGINEERING DIVISION, PENN STATE UNIVERSITY

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

SWENG 541: Advanced Database Design Concepts

Handling the Buffer Pool

ENGINEERING DIVISION, PENN STATE UNIVERSITY

Force every write to disk?

Poor

response time. But provides durability.

No Steal Force

Steal

Steal buffer-pool frames from uncommited Xacts?

If

Trivial

not, poor throughput. If so, how can we ensure atomicity?

No Force

Desired

SWENG 541: Advanced Database Design Concepts

More on Steal and Force

ENGINEERING DIVISION, PENN STATE UNIVERSITY

STEAL (why enforcing Atomicity is hard) To steal frame F: Current page in F (say P) is written to disk; some Xact holds lock on P.
What

if the Xact with the lock on P aborts? Must remember the old value of P at steal time (to support UNDOing the write to page P).

NO FORCE (why enforcing Durability is hard) What if system crashes before a modified page is written to disk? Write as little as possible, in a convenient place, at commit time,to support REDOing modifications.

SWENG 541: Advanced Database Design Concepts

Basic Idea: Logging

ENGINEERING DIVISION, PENN STATE UNIVERSITY

Record REDO and UNDO information, for every update, in a log.

Sequential

writes to log (put it on a separate disk). Minimal info (diff) written to log, so multiple updates fit in a single log page.

Log: An ordered list of REDO/UNDO actions

Log

record contains: <XID, pageID, offset, length, old data, new data> and additional control info (which well see soon).
7

SWENG 541: Advanced Database Design Concepts

Write-Ahead Logging (WAL)

ENGINEERING DIVISION, PENN STATE UNIVERSITY

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 Xact before commit.

#1 guarantees Atomicity. #2 guarantees Durability.

Exactly how is logging (and recovery!) done?

Well

study the ARIES algorithms.

8

SWENG 541: Advanced Database Design Concepts

WAL & the Log

ENGINEERING DIVISION, PENN STATE UNIVERSITY

DB LSNs pageLSNs

RAM flushedLSN

Each log record has a unique Log Sequence Log records flushed to disk 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.
pageLSN

System keeps track of flushedLSN.

The

max LSN flushed so far. flushedLSN

Log tail in RAM

WAL: Before a page is written,

pageLSN
9

SWENG 541: Advanced Database Design Concepts

Log Records
ENGINEERING DIVISION, PENN STATE UNIVERSITY

LogRecord fields:
prevLSN XID type pageID length offset before-image after-image

update records only

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

UNDO actions
10

SWENG 541: Advanced Database Design Concepts

Other Log-Related State

ENGINEERING DIVISION, PENN STATE UNIVERSITY

Transaction Table:
One

entry per active Xact. Contains XID, status (running/commited/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.
11

SWENG 541: Advanced Database Design Concepts

Normal Execution of an Xact

ENGINEERING DIVISION, PENN STATE UNIVERSITY

Series of reads & writes, followed by commit or abort.

We

will assume that write is atomic on disk.

practice, additional details to deal with non-atomic writes.

In

Strict 2PL.
STEAL, NO-FORCE buffer management, with Write-

Ahead Logging.
12

SWENG 541: Advanced Database Design Concepts

Checkpointing

ENGINEERING DIVISION, PENN STATE UNIVERSITY

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).

13

SWENG 541: Advanced Database Design Concepts

The Big Picture: Whats Stored Where

ENGINEERING DIVISION, PENN STATE UNIVERSITY

LOG LogRecords
prevLSN XID type pageID length offset before-image after-image

DB Data pages
each with a pageLSN

RAM Xact Table

lastLSN status

Dirty Page Table

recLSN

master record

flushedLSN
14

SWENG 541: Advanced Database Design Concepts

Simple Transaction Abort

ENGINEERING DIVISION, PENN STATE UNIVERSITY

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!

15

SWENG 541: Advanced Database Design Concepts

Abort, cont.
ENGINEERING DIVISION, PENN STATE UNIVERSITY

To perform UNDO, must have a lock on data!

No problem! You continue logging while you UNDO!! CLR has one extra field: undonextLSN

Before restoring old value of a page, write a CLR:

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!)
16

At end of UNDO, write an end log record.

SWENG 541: Advanced Database Design Concepts

Transaction Commit
ENGINEERING DIVISION, PENN STATE UNIVERSITY

Write commit record to log. All log records up to Xacts lastLSN are flushed.

that flushedLSN lastLSN. Note that log flushes are sequential, synchronous writes to disk. Many log records per log page.
Guarantees

Commit() returns. Write end record to log.

17

SWENG 541: Advanced Database Design Concepts

Crash Recovery: Big Picture

ENGINEERING DIVISION, PENN STATE UNIVERSITY

Oldest log rec. of Xact active at crash

Smallest recLSN in dirty page table after Analysis

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. u (repeat history) UNDO effects of failed Xacts.

Last chkpt CRASH

A R U

18

SWENG 541: Advanced Database Design Concepts

Recovery: The Analysis Phase

ENGINEERING DIVISION, PENN STATE UNIVERSITY

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 lastLSN=LSN, change Xact status on commit. Update record: If P not in Dirty Page Table, Add P to D.P.T., set its recLSN=LSN.

19

SWENG 541: Advanced Database Design Concepts

Recovery: The REDO Phase

ENGINEERING DIVISION, PENN STATE UNIVERSITY

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 > LSN, or pageLSN (in DB) LSN.

To REDO an action:
Reapply logged action. Set pageLSN to LSN. No additional logging!

20

SWENG 541: Advanced Database Design Concepts

Recovery: The UNDO Phase

ENGINEERING DIVISION, PENN STATE UNIVERSITY

ToUndo={ l | l a lastLSN of a loser Xact} Repeat:

Choose

largest LSN among ToUndo. If this LSN is a CLR and undonextLSN==NULL 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.

21

SWENG 541: Advanced Database Design Concepts

Example of Recovery
LSN
ENGINEERING DIVISION, PENN STATE UNIVERSITY

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
22

RAM
Xact Table lastLSN status Dirty Page Table recLSN flushedLSN

00
05 10 20 30 40 45

prevLSNs

50

ToUndo

60

SWENG 541: Advanced Database Design Concepts

Example: Crash During Restart!

LSN
00,05 10 20 Xact Table lastLSN status Dirty Page Table recLSN flushedLSN 30
ENGINEERING DIVISION, PENN STATE UNIVERSITY

begin_checkpoint, end_checkpoint update: T1 writes P5 update T2 writes P3 T1 abort update: T3 writes P1 update: T2 writes P5 CRASH, RESTART
undonextLSN

LOG

RAM

40,45
50 60

CLR: Undo T1 LSN 10, T1 End

ToUndo

70
80,85 90

CLR: Undo T2 LSN 60

CLR: Undo T3 LSN 50, T3 end CRASH, RESTART CLR: Undo T2 LSN 20, T2 end
23

SWENG 541: Advanced Database Design Concepts

Additional Crash Issues

ENGINEERING DIVISION, PENN STATE UNIVERSITY

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.

24

SWENG 541: Advanced Database Design Concepts

Summary
ENGINEERING DIVISION, PENN STATE UNIVERSITY

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.

25

SWENG 541: Advanced Database Design Concepts

Summary (Cont.)
ENGINEERING DIVISION, PENN STATE UNIVERSITY

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!

26

SWENG 541: Advanced Database Design Concepts

References
Lecture
ENGINEERING DIVISION, PENN STATE UNIVERSITY

notes based on Ramakrishnan, R. and Gehrke, J. (2003). Database Management Systems, Third Edition, McGraw-Hill, Boston, MA, pp. 579 602.

27