Swingbench

Hi,
I am trying to configure Swingbench for my test database but I am not able to configure connect string. Can anyone guide me to figure it out how to do ?
Assumed details.
DB ip : 10.10.10.25
SID = test
listener port = 1521
hostname = testserver
I am looking for oracle 10g type IV jdbc driver thin.

So where did you put the JDBC drivers?
Can we assume
You downloaded and unzipped the JDBC drivers in /oracle OR
You downloaded and unzipped the Instant Client in /oracle
Can we assume you updated the environment file as indicated in http://dominicgiles.com/installation.html

Similar Messages

SwingBench an Oracle Stress Test and Benchmark tool

Last week I was in a RAC workshop hosted by Oracle.
In the workshop, Oracle using a tool called Swingbench to generate load for RAC testing. Which is pretty interesting and good tools.
Since I have seen quite a few inquiry regarding a good benchmarking and load generating tools. I thought it might be good idea to share it.
It's free and can be downloaded here.
http://www.dominicgiles.com/swingbench.html
Swingbench is a free load generator (and benchmarks) designed to stress test an Oracle database (9i or 10g)

Hi,
To add to what yingkuan said, there is one more free stress testing tool called hammerora.
http://hammerora.sourceforge.net/
Regards

Swingbench error when starting swingbench. (more like java problem)

Platform: Solari 10, Oracle 10gR2.
I downloaded the swingbench for load testing from http://www.dominicgiles.com/swingbench.php for version 2.2 and 2.3. and unzip it in Solari 10 at $HOME (/export/home/oracle/swingbench). As result, it create a folder ./swingbench with all objects in it. I edited the file swingbench.env:
export JAVAHOME=/usr/java
export SWINGHOME=$HOME/swingbench
And leave the others as default. I followed the README.txt run the script swingbench.env to setup the env variables and get into $HOME/swingbench/bin to run ./swingbench to expect a GUI window. But I got this error:
fx1db01|inbound|/export/home/oracle/swingbench/bin $ ./swingbench
Exception in thread "main" java.lang.UnsupportedClassVersionError: com/dom/benchmarking/swingbench/SwingBench (Unsupported major.minor version 49.0)
at java.lang.ClassLoader.defineClass0(Native Method)
at java.lang.ClassLoader.defineClass(ClassLoader.java:495)
at java.security.SecureClassLoader.defineClass(SecureClassLoader.java:110)
at java.net.URLClassLoader.defineClass(URLClassLoader.java:251)
at java.net.URLClassLoader.access$1(URLClassLoader.java:217)
at java.net.URLClassLoader$1.run(URLClassLoader.java:198)
at java.security.AccessController.doPrivileged(Native Method)
at java.net.URLClassLoader.findClass(URLClassLoader.java:192)
at java.lang.ClassLoader.loadClass(ClassLoader.java:300)
at sun.misc.Launcher$AppClassLoader.loadClass(Launcher.java:290)
at java.lang.ClassLoader.loadClass(ClassLoader.java:256)
at java.lang.ClassLoader.loadClassInternal(ClassLoader.java:316)
What I can do to fix the problem?
Thanks for help!
Edited by: user5973955 on Aug 14, 2009 8:58 AM

You are trying to run a program compiled with target class version of 1.5 on a JRE version less than 1.5. Use JRE 1.5 or higher.

Swingbench Load Generator fails to quit the users

Hello All,
We have an Oracle DB 10.2 running on HP-UX. This is a demo application to demonstrate the migration of the CPU on the vPar environment when the oracle load goes high. We are using swingbench to generate the load to the Oracle DB.
The problem here is, the swingbench is configured with 7 users to generate the load to the Database. The load is not stable on the swingbench. It started with full load and it goes down to half. While analyzing the issue, we found that the "PL/SQL lock timer" is created on the db and fails to clear that lock. Basically it should fluctuates and it should goes to 0 sometimes. But in this case, it keeps on 7 locks. Due to this, the swingbench fails to logs off the users after the load generation completes.
Is this the problem with the db or related with the swingbench.
Please Advise.
Thanks,
Jayaprakash S

To Say in clear Words, We have a Oracle 10i Database and are using swingbench to increase the load in Oracle for an demonstration. But the swingbench is not maintaining the load on the oracle till the end of the scenario (10 mins).
The same setup is working for another instance of Oracle.

Swingbench don't show correct TPM?

Hi all,
i'm using swingbench for retrive TPM information, but i don't know why it shows me low TPM values.
My server is BL480C (2 CPU quad core) with 16Gb of RAM, oracle 10g SGA size is 6Gb and is sizing enough for OLTP.
i've create with ccwizard a schema of default 10000 customer and i've try a generation data for 100,1000,10000 and 40000 number of transactions.
When i test it with 25,100 or more users, in all case the result don't go over 5000 TPM for minute.. isn't possible! my server should be arrive at 80000 TPM
what i mistake in configuration of bench??
someone can help me please?
thanks !
Andrea

It's not clear if you think you have a problem with the configuration of your tool or you just can't believe that you can't drive your db to over 5k tpm.
If the first - configuration of your tool -- that's a question for your tool people, not Oracle.
If the second - perhaps you really can't drive your db - as currently hosted and configured - past 5k tpm. If you think it should go higher then you need to start analyzing your db performance just like any other tuning engagement.

Swingbench on RAC

Guys
I am trying to configure swingbench (ver 2.2) on 2 node RAC cluster. I am not able to make clusteroverview run. This is what I am doing..
1. start coordinator "./coordinator -g" -- Stars fine
2. Start swingbench "./swingbench"
3. ./clusteroverview
At step 3, I get error that "There is no client attached to coordinator"
4. ./coordinator -status
Coordinator is running and has 0 clients attached
Now the question was, How do I attach client to coordinator? Then in one option of swingbench command, I used following
./swingbench -co "//192.168.0.71/CoordinatorServer"
This does attach client to coordinator, but it my options to run load on swingbech screen is grayed out. I can't hit play button.
Any help/suggestion of this issue?
Thanks in advance
PG

Okay....
Apologies for the confusion... I thought I'd described this is in the 2.2 clusteroverview documentation
http://www.dominicgiles.com/swingbench/clusteroverview21f.pdf but looking at it is a little out of date
your nearly there...
the process is
1/ start the coordinator (say on node1)
2/ start the load generators specifying node 1 as the coordinator i.e.
./minibench -co //node1/CoordinatorServer -cs rac1 &
./minibench -co //node2/CoordinatorServer -cs rac2 &
This will start the load generators but will hand over control to the coordinator
3/ edit the clusteroverview.xml file to reflect your environment.
i.e
<?xml version='1.0'?>

<ClusterOverviewConfiguration>

<DisplayedCharts RefreshRate="3000" MaxPoints="1024">
<Chart>ControlPanel</Chart>
<Chart>TPM</Chart>
<Chart>UserConnections</Chart>
<Chart>Overview</Chart>
</DisplayedCharts>

<Coordinator>
<Location>//node1/CoordinatorServer</Location>
</Coordinator>

<MonitoredDatabaseList>
<MonitoredDatabase DisplayName="rac1" Username="system" Password="manager" DriverType="thin" ConnectString="node2:1521:RAC1" MonitoredUsers="CC,SOE"/>
<MonitoredDatabase DisplayName="rac2" Username="system" Password="manager" DriverType="thin" ConnectString="node3:1521:RAC2" MonitoredUsers="CC,SOE"/>
</MonitoredDatabaseList>
</ClusterOverviewConfiguration>
when clusteroverview starts you should be able to start the load generators from clusteroverview.... not from the loadgenerators themselves....
Does this help?
Dom

Large OLTP data set to get through the cache in our new ZS3-2 storage.

We recently purchased a ZS3-2 and are currently attempting to do performance testing. We are using various tools to simulate load within our Oracle VM 3.3.1 cluster of qty5 Dell m620 servers-- swingbench, vdbench, and dd. The OVM repositories are connecting via NFS. The Swingbench load testing servers have a base OS disk mounted from the repos and NFS mounts via NFS v4 from within the VM (we would also like to test dNFS later in our testing).
The problem I'm trying to get around is that the 256G of DRAM (and a portion of that for ARC) is large enough where my reads are not touching the 7200 RPM disks. I'd like to create a large enough data set so the amount of random reads cannot possible be stored within the ARC cache (NOTE: we have no L2ARC at the moment).
I've run something similar to this in the past, but have adjusted the "sizes=" to be larger than 50m. My thought here is that, if the ARC is up towards around 200 or so MB's, if I create the following on four separate VM's and run vdbench at just about the same time, it will be attempting to read more data than can possibly fit in the cache.
* 100% random, 70% read file I/O test.
hd=default
fsd=default,files=16,depth=2,width=3,sizes=(500m,30,1g,70)
fsd=fsd1,anchor=/vm1_nfs
fwd=fwd1,fsd=fsd*,fileio=random,xfersizes=4k,rdpct=70,threads=8
fwd=fwd2,fsd=fsd*,fileio=random,xfersizes=8k,rdpct=70,threads=8
fwd=fwd3,fsd=fsd*,fileio=random,xfersizes=16k,rdpct=70,threads=8
fwd=fwd4,fsd=fsd*,fileio=random,xfersizes=32k,rdpct=70,threads=8
fwd=fwd5,fsd=fsd*,fileio=random,xfersizes=64k,rdpct=70,threads=8
fwd=fwd6,fsd=fsd*,fileio=random,xfersizes=128k,rdpct=70,threads=8
fwd=fwd7,fsd=fsd*,fileio=random,xfersizes=256k,rdpct=70,threads=8
rd=rd1,fwd=fwd1,elapsed=900,interval=30,fwdrate=max,format=yes,pause=30,openflags=fsync
rd=rd2,fwd=fwd2,elapsed=900,interval=30,fwdrate=max,format=yes,pause=30,openflags=fsync
rd=rd3,fwd=fwd3,elapsed=900,interval=30,fwdrate=max,format=yes,pause=30,openflags=fsync
rd=rd4,fwd=fwd4,elapsed=900,interval=30,fwdrate=max,format=yes,pause=30,openflags=fsync
rd=rd5,fwd=fwd5,elapsed=900,interval=30,fwdrate=max,format=yes,pause=30,openflags=fsync
rd=rd6,fwd=fwd6,elapsed=900,interval=30,fwdrate=max,format=yes,pause=30,openflags=fsync
rd=rd7,fwd=fwd7,elapsed=900,interval=30,fwdrate=max,format=yes,pause=30,openflags=fsync
However, the problem I keep running into is that vdbench's java processes will throw exceptions
... <cut most of these stats. But suffice it to say that there were 4k, 8k, and 16k runs that happened before this...>
14:11:43.125 29 4915.3 1.58 10.4 10.0 69.9 3435.9 2.24 1479.4 0.07 53.69 23.12 76.80 16384 0.0 0.00 0.0 0.00 0.0 0.00 0.1 7.36 0.1 627.2 0.0 0.00 0.0 0.00 0.0 0.00
14:12:13.071 30 4117.8 1.88 10.0 9.66 69.8 2875.1 2.65 1242.7 0.11 44.92 19.42 64.34 16384 0.0 0.00 0.0 0.00 0.0 0.00 0.1 12.96 0.1 989.1 0.0 0.00 0.0 0.00 0.0 0.00
14:12:13.075 avg_2-30 5197.6 1.52 9.3 9.03 70.0 3637.8 2.14 1559.8 0.07 56.84 24.37 81.21 16383 0.0 0.00 0.0 0.00 0.0 0.00 0.1 6.76 0.1 731.4 0.0 0.00 0.0 0.00 0.0 0.00
14:12:15.388
14:12:15.388 Miscellaneous statistics:
14:12:15.388 (These statistics do not include activity between the last reported interval and shutdown.)
14:12:15.388 WRITE_OPENS Files opened for write activity: 89 0/sec
14:12:15.388 FILE_CLOSES Close requests: 81 0/sec
14:12:15.388
14:12:16.116 Vdbench execution completed successfully. Output directory: /oracle/zfs_tests/vdbench/output
java.lang.RuntimeException: Requested parameter file does not exist: param_file
at Vdb.common.failure(common.java:306)
at Vdb.Vdb_scan.parm_error(Vdb_scan.java:50)
at Vdb.Vdb_scan.Vdb_scan_read(Vdb_scan.java:67)
at Vdb.Vdbmain.main(Vdbmain.java:550)
So I know from reading other posts, that vdbench will do what you tell it (Henk brought that up). But based on this, I can't tell what I should do differently to the vdbench file to get around this error. Does anyone have advice for me?
Thanks,
Joe

ah... it's almost always the second set of eyes. Yes, it is run from a script. And I just looked and realized that the list last line didn't have the \# in it. Here's the line:
"Proceed to the "Test Setup" section, but do something like `while true; do ./vdbench -f param_file; done` so the tests just keep repeating."
I just added the hash to comment that out and am rerunning my script. My guess is that it'll complete Thanks Henk.

Stress Test question

Hi,
i've perform a stress test of a database in 2 configuration :
1) Single instance 11.2 with 8gb Sga e 4gb Pga on Linux, 6core with instance caging =4, resource_manager=defaut_plan
2) 2 Nodes Rac 11.2 with 8gb Sga e 4gb Pga in both nodes on Linux, 32core with instance caging =4, resource_manager=defaut_plan
I've used Swingbench and the transaction per minute are equal, Avg transaction per s in rac are Greater then single instance (4ms vs 40ms).
I know that rac is not for high performance.
My question is: is rac config best than single instance and if yes what is the %?
Thank you

Hi SBHV,
As far as I know, if you choose to test Audio Conferencing, it will also simulate Video conferencing.
And the Key Health Indicators in Lync Server 2013 for your reference.
http://technet.microsoft.com/en-us/library/dn593599.aspx#Front
Best regards,
Eric

Which Hardware shud i go? For this requirement.

Hello all,
As per requirement, In Oracle 9i, 1000 tps(transaction per sec) requirement ( 60% insert update and 40 % search operations). Dose oracle 9i ( standard ed.) with 2 CPU machine will handle this requirement ?
What will be performance figure on LINUX/SUN in term of tps in oracle 9i standard ed.?
Could you all please suggest me for what hardware platform should i go and also what RAM should i use here.
Database Size will be apporx 10G
Thanks in Advance!!
Its urgent please....
Pankaj

We have no idea what your hardware will be or whether these will be long running or short transactions or whatever.
Your question is not easy to answer; the only way of getting a definitive answer is to test your workload over a period of time.
For this there are tools which will load a system up : for example Hammer Ora or Swingbench are both freely available.

Services in RAC

Most of the time I feel uncomfortable when it come to creating service.
Can any one please share step by step procedure and more details on services and benefits around this.

Can any one please share step by step procedure and more details on services and benefits around thishttp://www.oracle.com/technology/obe/obe10gdb_vmware/ha/rac/rac.htm#t1
This tutorial covers the following topics:
Overview
Scenario
Prerequisites
Creating Services Using DBCA
Reviewing Your Enabled Services in Database Control
Using Swingbench to Generate Load on a RAC Database
Viewing Top Sessions to Evaluate Services
Simulating a Database Instance Crash
Summary
Related Information
Hope this helps you...

Swingbench2.4: Can not start schema creation wizards

Hi Community,
When trying swingbench2.4 I got "java.lang.NoClassDefFoundError: while resolving class: com.dom.benchmarking.swingbench.wizards.Wizard" error.
I've included $SWINGHOME/lib/swingbench.jar, where the wizard class file can be found, in my classpath, but that didn't work.
My JDK is jdk1.6.0_23.
Any help?
Thanks,
Todd

Hi Fahd,
Thanks, I've filled in all variables including below, but didn't work:
ORACLE_HOME
JAVAHOME
SWINGHOME,
ANTHOME=$SWINGHOME/lib
LD_LIBRARY_PATH=${LD_LIBRARY_PATH}:$ORACLE_HOME/lib
CLASSPATH=$JAVAHOME/jre/lib/rt.jar:$JAVAHOME/lib/tools.jar:$ORACLE_HOME/jdbc/lib/ojdbc6.jar:$SWINGHOME/lib/swingbench.jar:$ANTHOME/ant.jar
PATH=$JAVAHOME/bin:$PATH
Thanks

Oracle 11g, OL 6 and ext4

Now that Oracle Linux 6 has been certified for Oracle 11gR2, does this also include the use of the ext4 file system?
I'm asking because I found database performance and block corruption warnings searching in Google, but perhaps this only applied to early implementations of ext4 or were simply wrong assertions. I imagine that certification includes ext4, but just to be sure, is there still a reason to be concerned about it?

Dude wrote:
Concurrent I/O seems to be a topic often discussed when comparing XFS with ext4. If I'm not mistaken then ext4 supports concurrent I/O, but there is a performance penalty compared to XFS, i.e. fragmentation. However, I have not found any information outlining that this was a new issue introduced with ext4.
Apparently people are storing Oracle datafiles under ext3 without problems. The ext4 filesystem is the successor of ext3 and ext2, providing more features and robustness. How performance issues of Oracle DB pertain to ext4 is probably subject to errata and updates.
From what I can gather, ASM (ACFS), or OCFS2 in a RAC environment are the best options under Linux. Performance results will also depend on the underlying storage technology, for instance, caching, SSD or conventional rotary devices.It supports concurrent reads, not writes so if you're running a warehouse type database where it's mostly reads, you'd be fine but if you're doing a lot of writes, you should consider other options (assuming you care about latency and performance).
Personally, I think ASM is the way to go these days for Oracle DBs and you're right, there are a ton of factors to consider when architecting your ASM configs. For example, Oracle suggests putting all your database disks in one huge diskgroup because they claim that performance is better in that case, though I've run quite a lot of tests over the last couple years suggesting otherwise. At the end of the day, most people don't care about performance (though they may say otherwise) and certainly don't take the time to do much tuning. My situation was a bit different due to the nature of the business but even Jim Williams and some of the other folks on the ASM team didn't seem to have initially gone as in-depth as I may have when it came to tuning.
The best thing you can do is run your own tests with tools like fio, orion, swingbench, etc. and figure out if the performance is acceptable based on your needs.

Test oracle insert very very slow

I have just created a 11.2.0.3 db based on datawarehouse template, and tested insert 10,000,000 records through sysbench , and took 1 hour.
IT is extremely slow. I have not tune the db parameters yet, I assume the parameters are good for datawarehouse.
OS is redhat linux.
Anybody has advice on this?
Thanks in advance.

Having read [url http://sysbench.sourceforge.net/docs/]the manual, it sounds like it's doing row-by-rows inserts via a single session.
So not really a valid test.
How about:
Orion
[url http://www.dominicgiles.com/swingbench.html]Swingbench
[url http://kevinclosson.wordpress.com/2012/02/06/introducing-slob-the-silly-little-oracle-benchmark/]SLOB
These tools will help you benchmark and test your IO subsystems.
But, by far the most crucial element will be whether you insert 10,000,000 rows row-by-row or in bulk.
See Tim Gormans's Scaling to Infinity
not currently available from his website because it's being revamped but copies are out there, e.g.
http://www.nocoug.org/download/2012-11/NoCOUG_201211_Tim_Gorman_Data_Warehouse_Partitioning.pdf

Inputs to setup an Oracle performance tuning Lab.

Hi All,We are planning to setup a Performance tuning lab at our unit.The key areas which needs to be addresses for this lab is 1)Architecture of the Database 2)Having Large volume of data 3)Producing high volume transactions.
Request you to provide your valuable inputs on how to setup the lab.

user11266214 wrote:
Hi All
Thanks for your responses.We are not doing the load testing.We have a linux server on which Oracle 11g running.We would like to simulate the real time work environment on which we train people to perform the performance tuning tasks.Performance tuning tasks comprise of Server tuning,Oracle parameter tuning and SQL tuning.
Request you to provide inputs on the approach we need to take up for setting up such kind of lab.
Thanks in anticipation.Swingbench as said before

Which Operating system most company use to run oracle?

Which Operating system most company use to run oracle?
thanks

If you are trying to decide on a platform, I would recommend that you look at Swingbench and run your own benchmarks on various platforms using vendor "try and buy" options. Sun is great for allowing a 90 day "try and buy" where you could try a wide range of platforms and I am guessing other vendors would do the same. Using Swingbench I recently found that an older Sun X6220 (blade server) was for 50 Swingbench users faster than a T6340, T2000, M5000 and older V890. Single threaded SQL was especially slow on the SPARC/T while the M5000 was comparable to the X6220 (again for 50 users). This is significant because the M5000 is much more expensive than the X6220 (a blade server).
For a small number of users, say 50 users and under, I like Solaris/x86 unless there will be lots of IO which is where SPARC and the M series scales up very well. 50+ users hammering on a DB, Solaris/SPARC is my first choice. If you are just getting started and are not in love with any platform, check out Oracle Unbreakable Linux.

Swingbench

Similar Messages

Maybe you are looking for