Create your own awesome maps

Even on the go

with our free apps for iPhone, iPad and Android

Get Started

Already have an account?
Log In

SLOB by Mind Map: SLOB
0.0 stars - reviews range from 0 to 5


Test Models

Physical Read Performance

Use a small db_cache_size

In use specific parameters, alter system set "_db_block_prefetch_limit"=0, alter system set "_db_block_prefetch_quota"=0, alter system set "_db_file_noncontig_mblock_read_count"=0, alter system set compatible=""

Should see db file sequential read as top wait event

Redo Generation Performance

Use large redo logs

Use a large db_cache_size

alter system set log_checkpoint_timeout=99999999

Physical Write Performance

Use small redo logs

Use a small db_cache_size, With small db_cache_size, don't need to include readers, will have physical reads

Logical I/O Performance

use a big db_cache_size (bigger than tables used)

avoid any auto memory management

Analysis Tools

Included in SLOB distribution

Summarizes multiple AWR reports

SLOB.R, Download,, Interrogates output in R, Requires R Statistical Environment, Download,


General Setup

Eliminate other activity on database server

Disable automatic AWR snapshots

alter system set resource_manager_plan=''

Disable parallel unless testing parallel performance

alter system set parallel_max_servers=0

alter system set cpu_count=1

Switch logfiles (N*2+1) times before each run

N = number of redo logs

Do a throwaway SLOB run after a bounce to prime the database

Save awr.txt output after each run for further analysis

Scripts to automate multiple SLOB runs and save awr.txt

Try misc/create_database_kit/ in SLOB to build a testing DB

Isolate UNDO blocks from table/index (chains) for write/REDO model by setting db_recycle_cache_size=N


Introducing SLOB:

Collected links/tips/tools:

Yury Velikanov's SLOB index:

SLOB LinkedIn Group:

Karl Arao's TiddlyWiki:[[cpu%20-%20SillyLittleBenchmark%20-%20SLOB]]


Benchmark one storage unit vs another

Compare performance of various filesystems

Don't mix LUNs; use same LUN for each FS

Identify max throughput possible from your setup

Non-SLOB tools




Other Considerations

Distributing your SLOB results/AWRs may be a legal issue with Oracle Corp

Scale data against a baseline?