Performance
Joseph Meeks
Director, Product Management
Oracle High Availability Systems
Note to viewer
Note to viewer
See this FAQ for more customer and sales collateral
http://database.us.oracle.com/pls/htmldb/f?
p=301:75:101451461043366::::P75_ID,P75_AREAID:21704,2
43 seconds
2,000 sessions
on both primary
and standby
P
5Copyright 2012, Oracle and/or its affiliates. All rights reserved.
n
i
m
i
l
re
# of database
sessions on
primary and
y
r
a
48 seconds
2,000 sessions
on both primary
and standby
# of database
sessions on
primary and
y
r
a
72 seconds
83 seconds
1,000 sessions on
both primary and
standby
500 sessions on
both primary and
standby
P
6Copyright 2012, Oracle and/or its affiliates. All rights reserved.
n
i
m
i
l
re
# of database
sessions on
primary and
standby
# of
database
sessions on
primary and
The following performance tests show the impact of SYNC transport on primary
Database 12c (same process as above, but standby acks primary as soon as
redo is received in memory it does not wait for SRL write.
8Copyright 2012, Oracle and/or its affiliates. All rights reserved.
OLTP with Random Small Insert < 1ms RTT Network Latency
Workload:
Random small inserts (OLTP) to 9 tables with 787 commits per second
132 K redo size, 1368 logical reads, 692 block changes per transaction
OLTP with Random Small Inserts and < 1ms RTT Network Latency
Local standby,
104,051,368.80
Txn Rate
20000000
40000000
60000000
80000000
100000000
<1ms RTT
99MB/s redo rate
<1% impact on
database
throughput
1% impact on
transaction rate
120000000
Swingbench OLTP
6000
5000
4000
6363
6151
6077
tps
tps
tps
Baseline
No Data Guard
3000
2000
1000
0
30 MB/s redo
3% impact at
1ms RTT
5% impact at
5ms RTT
Transaction profile
440K redo size, 6000 logical reads, 2100 block changes per transaction
200
150
83 MB/s redo
<1%% impact
189
188
177
167
tps
tps
tps
tps
2ms RTT
Network
Latency
5ms RTT
Network
Latency
100
50
0
Baseline
No
Data Guard
1ms RTT
Network
Latency
at 1ms RTT
7% impact at
2ms RTT
12% impact at
5ms RTT
Transaction profile
4K redo size, 51 logical reads, 22 block changes per transaction
30,000
25,000
Txn Rate
Redo Rate
20,000
15,000
10,000
5,000
0
No Sync
0ms
2ms
5ms
10ms
20ms
Agenda
Failover and Switchover Timings
SYNC Transport Performance
ASYNC Transport Performance
Primary Performance with Multiple Standby Databases
Redo Transport Compression
Standby Apply Performance
transmission is completed
ASYNC has minimal impact on primary database performance
Network latency has little, if any, impact on transport throughput
Uses Data Guard 11g streaming protocol & correctly sized TCP send/receive buffers
Performance tests are useful to characterize max redo volume that ASYNC is
600
500
Redo
Transport
MB/sec
400
high rates
484
300
200
100
0
Single Instance
or more instances
23Copyright 2012, Oracle and/or its affiliates. All rights reserved.
Redo
Transport
Rate
MB/sec
35
30
25
20
15
Network
Latency
10
0ms
25m
s
5
0
ASYNC
X$LOGBUF_READHIST to determine
Agenda
Failover and Switchover Timings
SYNC Transport Performance
ASYNC Transport Performance
Primary Performance with Multiple Standby Databases
Redo Transport Compression
Standby Apply Performance
Multi-Standby Configuration
Primary - A
Local Standby - B
Guard configurations.
SYNC
ASYNC
(compared to baseline)
(compared to baseline)
105.0%
104.0%
103.0%
102.0%
101.0%
100.0%
99.0%
98.0%
97.0%
0 - 14 destinations
27Copyright 2012, Oracle and/or its affiliates. All rights reserved.
102.0%
101.0%
100.0%
99.0%
98.0%
97.0%
96.0%
95.0%
94.0%
93.0%
0 -14 destinations
104.0%
102.0%
(compared to baseline)
100.0%
102.0%
98.0%
100.0%
96.0%
98.0%
96.0%
94.0%
Zero
1/0
1/1
1/14
# of SYNC/ASYNC destinations
28Copyright 2012, Oracle and/or its affiliates. All rights reserved.
92.0%
Zero
1/0
1/1
1/14
# of SYNC/ASYNC destinations
database
29Copyright 2012, Oracle and/or its affiliates. All rights reserved.
Agenda
Failover and Switchover Timings
SYNC Transport Performance
ASYNC Transport Performance
Primary Performance with Multiple Standby Databases
Redo Transport Compression
Standby Apply Performance
Transport
Lag - MB
22 MB/sec
uncompressed
available bandwidth
Recovery Point Objective (RPO)
impossible to achieve
perpetual increase in transport lag
12 MB/sec
compressed
Agenda
Failover and Switchover Timings
SYNC Transport Performance
ASYNC Transport Performance
Primary Performance with Multiple Standby Databases
Redo Transport Compression
Standby Apply Performance
cache
On Exadata x2-2 quarter rack
Swing bench OLTP workload
72 MB/second apply rate
I/O bound during checkpoints
1,762ms for checkpoint
complete
110ms DB File Parallel Write
633ms vs 1,762ms
DB File Parallel Write is
21ms vs 110ms
35Copyright 2012, Oracle and/or its affiliates. All rights reserved.
Allstate Insurance
Data Warehouse ETL processing resulted in average apply rate over a 3
Standby
Apply
Rate
700
600
500
400
300
200
100
0
MB/sec