VMware Communities : All Content - VMware VMmark™

error durring javaserver workload result collection

simonpha — Sat, 21 Nov 2009 04:16:28 GMT

Hi I've run into this problem with the javaserver workload when completing the workload run. I receive the following error in STAX. It seems to me that the workload completes correctly but when trying to do the result collection it errors out.

20091119-18:55:02 Start JobID: 2, File: C:\vmmark\xml\vmmark_main.xml, Machine: local://local, Function: Main, Args: { 'CONFIGFILE' : r'C:\vmmark\VMMARK.CONFIG' }, JobName: smalltest1
20091119-18:55:02 Info Holding block: main
20091119-18:55:02 Info Received RELEASE BLOCK main request
20091119-18:55:02 Info Releasing block: main
20091119-18:55:13 Start Testcase: VMmark Setup
20091119-18:55:36 Stop Testcase: VMmark Setup, ElapsedTime: 00:00:23
20091119-18:55:36 Start Testcase: VMmark Wkld Setup
20091119-18:57:54 Stop Testcase: VMmark Wkld Setup, ElapsedTime: 00:02:18
20091119-18:57:54 Start Testcase: VMmark Run Tiles
20091119-19:53:24 Stop Testcase: VMmark Run Tiles, ElapsedTime: 00:55:30
20091119-19:53:24 Start Testcase: VMmark Results Collection
20091119-19:53:25 Error STAXPythonEvaluationError signal raised. Terminating job.

===== XML Information =====

File: c:\vmmark\xml\javaserver_functions.xml, Machine: local
Line 322: Error in element type "call".

===== Python Error Information =====

com.ibm.staf.service.stax.STAXPythonEvaluationException:

Python object evaluation failed for:
'Info: COPY FILE C:\javaserver_%u.stdout TOFILE %s%sjavaserver_%u.stdout' % (tileserver, gResultsDir[0], gFsep, tileserver)

Traceback (innermost last):
File "<pyEval string>", line 1, in ?
ValueError: invalid literal for __int__: javaserver0
===== Call Stack for STAX Thread 1 =====

[
function: Main (Line: 106, File: C:\vmmark\xml\vmmark_main.xml, Machine: local://local)
sequence: 16/22 (Line: 128, File: C:\vmmark\xml\vmmark_main.xml, Machine: local://local)
testcase: VMmark Results Collection (Line: 507, File: C:\vmmark\xml\vmmark_main.xml, Machine: local://local)
sequence: 1/2 (Line: 508, File: C:\vmmark\xml\vmmark_main.xml, Machine: local://local)
iterate: 1/5 0 tilelist (Line: 510, File: C:\vmmark\xml\vmmark_main.xml, Machine: local://local)
sequence: 1/1 (Line: 511, File: C:\vmmark\xml\vmmark_main.xml, Machine: local://local)
iterate: 2/6 JavaServer gWORKLOADLIST (Line: 513, File: C:\vmmark\xml\vmmark_main.xml, Machine: local://local)
sequence: 2/2 (Line: 514, File: C:\vmmark\xml\vmmark_main.xml, Machine: local://local)
function: GetResultsJavaServer (Line: 227, File: c:\vmmark\xml\javaserver_functions.xml, Machine: local)
sequence: 8/12 (Line: 234, File: c:\vmmark\xml\javaserver_functions.xml, Machine: local)
]
20091119-19:53:25 Info Terminating block: main
20091119-19:53:25 Stop Testcase: VMmark Results Collection, ElapsedTime: 00:00:01
20091119-19:53:25 Status Testcase: VMmark Run Tiles, Pass: 1, Fail: 0, ElapsedTime: 00:55:30, NumStarts: 1
20091119-19:53:25 Status Testcase: VMmark Setup, Pass: 1, Fail: 0, ElapsedTime: 00:00:23, NumStarts: 1
20091119-19:53:25 Status Testcase: VMmark Wkld Setup, Pass: 1, Fail: 0, ElapsedTime: 00:02:18, NumStarts: 1
20091119-19:53:25 Status Testcase Totals: Tests: 3, Pass: 3, Fail: 0
20091119-19:53:25 Status Job Result: None
20091119-19:53:26 Stop JobID: 2

Anyone have any idea what this may be?

Thanks in advance

Tool to measure VMware performance or VMmark

sudiady — Wed, 18 Nov 2009 12:07:16 GMT

Look like VMmark is the current and only industry standard, it's relatively fair to use the same tool for other hypervisor products (ex. XenServer & etc)

Understanding the VMmark Results

lroderic — Tue, 17 Nov 2009 17:23:34 GMT

Hello. In Chapter 1 of the VMmark Benchmarking Guide, the section titled "VMmark Benchmark Scoring Methodology" describes the methodology for scoring. Table 1-3 shows the metrics for each workload. From the guide:

After a valid run, the metrics of the workloads within each tile are computed and aggregated into a score for that tile. This aggregation is performed by first normalizing the different performance metrics (such as MB/s and database commits/s) with respect to a reference system. Then a geometric mean of the normalized scores is computed as the final score for the tile.

The resulting per-tile scores are then summed to create the final metric. Normalization allows the integration of the different component metrics into an overall score. A reference system is commonly used for normalization when computing benchmark scores. For example, SPEC CPU2000 (see http://www.spec.org/cpu2000/) takes this approach, using a Sun Ultra5_10 with a 300Mhz processor as its reference platform. VMmark measures consolidation workloads within virtual environments and therefore requires a reference system capable of successfully running a single tile.

Use esxtop -b and output to a CSV file to see the system resource utliization during the benchmark run. Then analyze the data with Perfmon, Excel, or esxplot at http://vpivot.com/2009/10/21/esxtop-analysis-with-esxplot/. You'll want to see if I/O is balanced across HBAs. If not, you might consider changing the Path Selection method to Round Robin. Check the vSphere Performance white papers to make sure your HBA queue setting is optimal.

Use the performance monitoring utilities that the SAN provides as well. Though I/O might look fine from vSphere's perspective, the LUN configuration and placement of the workload datafiles can make a difference.

Since you're using VMmark for testing only and not for publication, you don't need to worry about compliance, and you're welcome to change the workloads. So if you find that VMmark doesn't drive enough disk I/O, you can, for example, adjust the fileserver workload to be heavier.

Lisa

Warning: RMI exception trying to contact client0:1098.

eMax04 — Tue, 17 Nov 2009 14:25:39 GMT

Somehow... when I changed over to the Gigabit switch in our lab, all these errors went away....

VCB Configuration Problem

eMax04 — Thu, 12 Nov 2009 14:01:29 GMT

Hey, I dont think your going to find an answer where you posted your problem.... you posted in the VMmark benchmark fourm....

Sorry dude!

2 ESX shows Red

weinstein5 — Thu, 04 Jun 2009 15:46:24 GMT

You might try removing the inactive ESX hosts form the inventory and readding them -

If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful

Is there a VMark Dependancy check?

psmith2006 — Thu, 05 Nov 2009 21:52:42 GMT

In the VMMARK.CONFIG you can try setting RUNFLAG as described in the comments to go thru the phases stepwise. That may help determine problems in the initial setup. Running single workload tests are also useful when debugging the initial setup. Reducing the RUNTIME during debugging is also useful. Please also review the troubleshooting section in the VMmark Benchmarking Guide .

Guest in Vm instance

weinstein5 — Thu, 05 Nov 2009 14:46:10 GMT

If vmware tools are installed - shutdown guest and restart guest will cause the VM to shutdown and restart respectvely through the operating system whether it is Linux or windows - power off guest or reset guest does a hard power off or hard reset of the vm -

If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful

SPECweb Wont install on Webserver Template

eMax04 — Wed, 04 Nov 2009 20:01:26 GMT

AWESOME!! it worked!! THANKS!

VMmark Client - Exchange Connectivity Errors

eMax04 — Wed, 04 Nov 2009 19:09:37 GMT

It looks like the original mailbox installed by Exchange was corrupt somehow. I uninstalled exchange and deleted all the files, then reinstalled it and now everything is working.

thanks for you help!!

Password and Username for the downloadable VM's

psmith2006 — Tue, 03 Nov 2009 16:14:23 GMT

Try: vmmark

Hi Guys.. my VMware get conrinuous reboot

kurva1979 — Tue, 27 Oct 2009 05:37:38 GMT

Hi Friends,

I am using vmware workstation and installed solaris 10, every thing look good for 1 month and later it get loop reboop( it get reboot continuour at GRUB options after 10 seconds) you can find the attachment .. pls can any one help me how to resolve this issue.. as it is very important for me pls guys.. help me to come out from this issue.. tks in Advance...

I keep getting an error message "Error Caused by File"

kevinadavis — Sat, 24 Oct 2009 08:27:47 GMT

I am brand new to the virtual world and just recently installed blade servers and VMWare. I am running VSphere 4.0. I created my first VM Template and when I try and create a new VM from that template I get an error message "Error Caused by File" and the VM creation terminates. I created a 600 GB volume on my SAN to place the VMs I create so space should not be an issues since I only have my virtual center VM currently in that volume. Any help will be greatly appreciated.

Thank You!

Kevin

VMware dead

weinstein5 — Fri, 09 Oct 2009 14:04:16 GMT

Welcome to the Forums - What product are you using?

If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful

clone_vm2.sh

CSS31 — Wed, 07 Oct 2009 13:57:43 GMT

rename_vm.sh

CSS31 — Wed, 07 Oct 2009 14:00:46 GMT

webserver workload failing with "[FATAL] Exceeded max allowed overthink time" errors

jamesz08 — Fri, 02 Oct 2009 20:40:06 GMT

Are you using the VMmark template for the webserver? Are the other workloads running fine when you do an entire tile? I have only seen this issue when there is a config problem or the system is extremely overloaded.

STAF failing to start in webserver and fileserver systems

qc4vmware — Wed, 30 Sep 2009 23:07:05 GMT

OK... I did indeed screw up something when re-ip'ing. I updated one hosts file and copied it to the other two nodes and forgot to update the local host entry. DOH!

What domain name I should add to the client system?

lroderic — Wed, 23 Sep 2009 14:02:07 GMT

Tobias is correct. See p. 91 in the Benchmarking Guide 1.1 for promoting mailserver0.mycompany.com to be a domain server. Then once mailserver0.mycompany.com is its own domain, add client0 to the mailserver0.mycompany.com domain so that client0's fully qualified name becomes client0.mailserver0.mycompany.com.

Lisa

Insatllation VMmark and pre requistions

creddya — Wed, 23 Sep 2009 05:06:28 GMT

Hi every one,

Can any body tell me where i have to install this in ESX Server or any physical machine i can use to create the set up

And tell me setup procedure..

Waiting for ur reply..

Vmark benchmark tool

creddya — Tue, 08 Sep 2009 05:35:17 GMT

Thanks for ur help ...Smith...

i have solved problem ....

thank u..

Running the client as a vm and yet more on individual workloads failing

HughBorg707 — Thu, 27 Aug 2009 23:38:21 GMT

Some of the things I would try would be to check to see if all VMs are using all 8 cores. If they are, try assigning some VMs to only use specific cores and others VMs the remaining ones.

Also if you don't have resource pools setup, try that as well with hard defined resource requirements. That way if a machine starts to misbehave, it will be limited to what the resource pool has available, not what the entire box has.

Beyond that, are all VMs using the same NICs etc?

If you try all this out, while things are running check the performance at all 3 levels. (box, resource pool, individual VM). That should help isolate where the problem lies. At the very least you'll have a better idea of what kind of resources your devices are needing and what you can do to correct the situation.

While not super technical answer, hopefully this does point you in an unexplored direction.

Regards

wiutyr

mmd7777 — Mon, 24 Aug 2009 14:35:20 GMT

fuy9uyer98

2. What is the default permission when a partition is formatted with NTFS

bazlina — Wed, 05 Aug 2009 15:20:49 GMT

1. What is the default permission when a partition is formatted with NTFS

2. What is the best way to secure files and folders that i share on NTFS partitions?

Upgrade Ram to 4Gb on Host But still only see 1.99Gb

DLeid — Fri, 31 Jul 2009 16:38:42 GMT

Okay, so we need to find out why the host doesn't see the memory. Since it was the last thing that was changed, try reseating the modules and probably substitute the memory to determine if it related to the new RAM modules you installed.

If you find this or any other information helpful or correct, please consider awarding points.

VMMark Nehalem Benchmark - effect of memory configuration

LucasAlbers — Thu, 23 Jul 2009 03:15:02 GMT

http://communities.vmware.com/message/1217680#1217680

800MHz clock speed.. Using 1333MHz memory (where supported) versus 1066MHz offers up to

*
9% better performance, while 1066MHz memory produces up to 28% better performance than

*

800MHz memory

ESX 3.5 vs. 4.0

LucasAlbers — Thu, 23 Jul 2009 03:11:24 GMT

http://it.anandtech.com/IT/showdoc.aspx?i=3560&p=4

AMD told us they expect that the VMmark scores of
the Opterons will be 5 to 10% higher on ESX 4.0. Although the Quad
socket numbers do not reflect this, it is a plausible claim. ESX 4.0
takes into account the processor cache architecture to optimize CPU
usage and has a finer grained locking mechanism. These improvements
should benefit all CPUs.

http://it.anandtech.com/IT/showdoc.aspx?i=3560&p=3

The new Xeon 5570 is between 55% and 157% faster,
depending on your source. Let's try to make sense out of this. First,
take the lowest score,

esx 3.5u4->esx 4.0

14.22. -> 23.50

Does that seem right?

DNS and VMmark

tbjs — Mon, 20 Jul 2009 16:59:57 GMT

Thanks James. I actually populated the host file last week and was able to get this working, just had no time to reply.

VMmark result

jamesz08 — Mon, 20 Jul 2009 16:59:55 GMT

There is an explanation of the scoring hte Benchmarking guide. If there are any parts of that which you have question on please reply to this.

The workloads are capped, so you cannot actually get 20@10 tiles. The maximum score for a single tile is ~1.54, there is small fluctuation in the scores.

ping v pc from other pc on the network

prakashraj — Fri, 10 Jul 2009 09:38:14 GMT

Hi ,

Are you able to ping the host machine from other machines on the network?

provide the IP configuraion details of your setup

Prakash

VMmark Run Hanging

iastadams — Thu, 09 Jul 2009 23:20:47 GMT

I finally found the resolution to this issue. One of the client servers did not have the DEP setting set to alwaysoff which was causing LoadSim to crash. Once the setting was changed then LoadSim ran without crashing.

VMmark and benchmarking of Hyper V and XEN

jamesz08 — Tue, 07 Jul 2009 18:33:52 GMT

Apendix F of the VMmark Benchmarking Guide contains information on cloning.

If you have Virtual Center you can clone through the UI.

Cloning through the CLI is not difficult, however it is a bit more involved since you will need to manualy modify each VM to give it a unique IP and hostname.

Either method will require the use of sysprep to clone the Windows systems. Sysprep is available on the Win2k3 CD in \support\tools\deploy.cab

The mailserver VMs must be cloned BEFORE making them a domain controller and installing Exchange.

VMmark 1.1.1 updated

lroderic — Wed, 17 Jun 2009 20:38:57 GMT

As documented for 1.1.1, attached is a sample script for creating passwordless ssh login between the client and ESX host.

The script's owner is Ludwik C. Siadlak (http://ludwikc.net), but it appears his site is off the net.

Lisa

Problems with file server workload

iastadams — Thu, 11 Jun 2009 17:12:07 GMT

That was it. Thanks for the help.

Questionable result in VMmark webserver test

Theovd — Mon, 08 Jun 2009 07:43:45 GMT

Hi all,

I found the issue: My test client was set to the Dutch locale and this should be U.S.

Theo.

Vmark

Rohail2004 — Fri, 29 May 2009 18:16:00 GMT

anyone using it? how does it work? help would be greatly apperciated

New Intel 5500 series processor

gary1012 — Wed, 20 May 2009 20:22:04 GMT

All of the top 8 core VMmark performers are 55x0 processors. Don't know if that helps you out...

vmware & vista

weinstein5 — Mon, 18 May 2009 18:44:02 GMT

Welcome to the Forums - is it Vista asking for username and password? Vista can be configured with user accounts requiring authentication against local user credentials -

If you find this or any other answer useful please consider awarding points by marking the answer correct or helpful

TEST DL785 32 Core + 128 Go RAM

psmith2006 — Thu, 07 May 2009 14:50:59 GMT

I would recommend that you download the VMmark kit and in it you will find the VMmark Benchmarking Guide, it explains how to configure the clients, server, and the VMs and the communication requirements. You can also download the VMmark appliances for the 3 Linux VMs used. The Windows VMs we can not provide, so you'll need to build those according to the directions in the Benchmarking Guide. You can use cloning for most of the VMs, except the mailserver VM which has special build requirements described in the guide.

As for the System and Storage configuration, I would recommend downloading the VMmark Full Disclosure Reports (FDR) for some that are similar to what you plan on testing. There should be enough info in the FDR to allow someone to reproduce a result within a few percent. Of course to get the same results will also require the same basic hardware. You however may have limitations on what hardware you have available, so you may not be able to get your system to the same level if for example - you don't have the same amount of RAM or your storage is older and not as fast or has less bandwidth than that used in the published VMmark results.

A twenty tile test is a lot of work to stage to get a fully compliant VMmark result, especially if you've never setup the benchmark for 1 or 2 tiles before.

SPECjbb2005 1.0.7 setup.jar: Could not load winAMD64ppk.dll

neilst — Thu, 23 Apr 2009 15:16:07 GMT

I had the same problem yesterday. To fix this I did the following:

Removed JRocket
Install 32bit SUN JRE (ver 6)
Installed SPECjbb2005 (all installed correctly)
Removed SUN jre
Installed JRocket

Regards

Neil

HP ProLiant DL785 G5 achieves #1 virtualization performance result on VMmark benchmark ( http://www.vmware.com/products/vmmark/results.html )

aaklea — Wed, 22 Apr 2009 01:26:52 GMT

HP ProLiant DL785 G5 achieves #1 virtualization performance result on VMmark benchmark

( http://www.vmware.com/products/vmmark/results.html )

Duration of a VMmark Run?

herevirtually — Fri, 10 Apr 2009 17:07:59 GMT

Thanks to psmith2006 for the reply and information ... I've marked the previous post as the answer to this thread.

Planning to move VMmark Community under VMTN > General > Performance

Badsah — Thu, 09 Apr 2009 00:55:41 GMT

After
receiving feedback from the community, we are planning to move the
VMmark community under the Performance community, as a sub-community.
This is a heads-up, although we will set up a redirect.

Thanks,
Badsah

---
Badsah Mukherji
VMware Web Communities Team

What about a VMware Storage Benchmark for Several Storage Vendors ?

meistermn — Mon, 09 Mar 2009 11:06:32 GMT

I like the VMMARK Benchmark for Servers, now will there be VMware Benchmark for Storage Systems .

This would be great. The Spec results for physical Storage System Benchmarks are not helpful.

where can I find the BEA JRockit?

psmith2006 — Wed, 04 Mar 2009 14:17:39 GMT

Correction: JRockit R27.5.0 for Java SE 6 for Microsoft Windows (64-bit x86)

Should be: JRockit R27.5.0 for Java SE 5 for Microsoft Windows (64-bit x86)

As of 03/04/2009 Current version is:

jrockit-R27.5.0-jdk1.5.0_14-windows-x64

Running / Troubleshooting the test harness ?

lroderic — Thu, 12 Feb 2009 20:28:19 GMT

Let's take this offline and handle through email.

Lisa

WebServer Template?

jamesz08 — Tue, 10 Feb 2009 22:28:40 GMT

You can boot into single user mode to get around this (add the word single to the boot options at the bootup screen). This will skip the network configuration. You can then setup your static IP and reboot.

setting MANDATORY_DEVICES=" " (that is a space) in /etc/sysconfig/network/config will disable the wait for mandatory devices. Or you can change the WAIT_FOR_INTERFACES entry to something small.

Cygwin environment install - 'Install to local directory' option must be selected

webcrew — Thu, 22 Jan 2009 15:41:04 GMT

If anything could be done to 'finish my comment' it would be a little more clarity in the 'Benchmarking Guide', as the cygwin site isn't clear on this

Thanks Lisa!

Tilescore.pl and SPECweb2005 v1.20

jamesz08 — Wed, 21 Jan 2009 22:01:56 GMT

The harness contains the sorce and a binary under the vclient/filesever/dbench/src directory. The tbench_srv.exe file is the windows binary that runs on the client system. The dbench and lock_mem files are Linux binaries that run on the fileserver VM.

Where are the new vmmark results for sun servers?

meistermn — Sat, 17 Jan 2009 00:30:53 GMT

Thanks for your Info.

SUN has updated the vmmark results for the Intel Rack Server X4450 with 24 Cores (Dunntington) on 13. Jan. 2009. Thanks for this.

You are right , the X4600 M2 with Shanghai is a beast. But I wait for X4600 M3 , when the hops problem is solved for all AMD CPU's with Hypertransport 3 and CPU Istanbul.

What happens, if you use all dimms Slot in the X4440 . Does the dimm speed drop from 667 to 533 MHZ? Or are 800 MHZ Dimms used (6400). Will it then drop from 800 to 667 MHZ?

In this thread is more about dimm speed drop.

http://communities.vmware.com/thread/51944

4 tiles got 4.91 scores but 5 tiles got 5.25 scores?

bherndon — Fri, 19 Dec 2008 05:52:48 GMT

Do you have CPU, network, and disk utilization data for 1, 2, 3, 4, and 5 tiles? It might be useful to help determine the bottleneck. I tend to use esxtop in batch mode.

Another way to look for bottlenecks is to compare your workload component scores with some of the ones published on the VMmark results page. A well-tuned tile should achieve an overall score 1.5-1.6 until CPU saturation. Most published results go a bit past saturation and the individual tiles scores are closer to 1.4. In your test, 4 tiles should achieve close to a score of 6 unless there is a bottleneck. That implies that at 4 tiles you already have a performance problem. I would suggest looking at each workload's normalized score and compare them with published numbers. That might tell you where to look. For instance, poor fileserver scores relative might point to storage.

If you post the tilescore.out, I can take a look.

Also, how is the storage configured - # of LUNs, RAID level, etc.?

Where can download VMDq(NETqueue)?

jamesz08 — Mon, 20 Oct 2008 17:25:46 GMT

Netqueue is supported on ESX 3.5 update 1 for Intel and Neterion 10Gbe NICs.

Please see the following KB article for details on how to enable this support:

http://kb.vmware.com/kb/1006272

Where's SPECweb_Ecommerce.Unix-PHP.config?

jamesz08 — Mon, 20 Oct 2008 16:53:01 GMT

That version is definitely the problem. You need a minimum of 1.10. VMmark 1.1 has been transition to SPECweb 1.20 which is the current release. Please contact SPEC to get the correct version.

james

VMmark - Appliances?

lroderic — Thu, 16 Oct 2008 14:23:04 GMT

Hi. We appreciate your feedback and work hard to make this benchmark as easy to deploy as we can.

Since the standby server runs none of the actual benchmarks, you probably can build your own with your own license faster than you can download an appliance. The same is true for the javaserver, and you need to purchase and manually install SPECjbb (since SPEC rules say we cannot redistribute).

Creating the mailserver requires domain promotion and Exchange installation, things you manually need to do with every mailserver in your testbed. We've researched this and asked Microsoft directly, and as far as we and they know there isn't a way to clone and reconfigure an Exchange server.

Please review the installation guide for instructions on creating template VMs that you then can clone using Virtual Center. Once you create template standby and javaservers, it's a matter of cloning, adjusting STAF, and changing the hostname/IP address. You can create a mailserver base VM (without domain promotion and Exchange installation), but the mailserver VM will always require manual installation.

Lisa

How to stop STAF(STAX) smoothly

psmith2006 — Mon, 13 Oct 2008 12:01:05 GMT

Check the troubleshooting section in the VMmark Benchmarking Guide around page 126.

Try the proceedure that talks about canceltest.sh

tgz Uncompress error

ceebee2 — Wed, 24 Sep 2008 07:00:10 GMT

Hello @all,

Problem solved: My mistake was to rely on graphical interfaces ... 7zip and Ark both stopped uncompressing the files with an error message saying that the archive was not valid. So I tried to untar them via the linux command shell with "tar xzvf filename.tgz" which worked perfectly fine. Thanks to fireguy for your hint! It seems that the graphical tools have some problems with large files.

Thanks @all for your help!

Execute STAX has an error message

clemenslee — Fri, 11 Jul 2008 08:12:40 GMT

Find result by myself

In STAF.cfg file, excute path should not erase.

workload hangs at "mailserverclient"

Notti — Mon, 19 May 2008 08:48:47 GMT

Sorry for my late response....

If your problem still exists:
1. log onto both client machines via RDP and the appropriate Domain Administrator for each Domain.

2. the quickest way to run STAXMon via local admin is to open a cmd-box, browse to your STAXMon Folder and type: runas /user:Mailserver00\administrator STAXMon (please replace administrator with your set local admin account) type in your local administrator password.... this should do the job!

I hope this helps out.... regards - Notti

workload hangs at "mailserverclient"

dsynnott — Tue, 06 May 2008 16:43:55 GMT

Jamesz08;

Thank you for your assistance. I made two changes, both with the scheduled tasks. (I did not have to re-install anything using the Domain Administrator login. The installation using the local administrator account workes fine.)

Changes that enabled the test to run:

1. Modified 'Run as' to Domain Administrator, i.e. MAILSERVER00\Administrator.

Start -> Control Panel -> Scheduled Tasks -> Start STAF 3.2

Initial Panel, modify 'Run As:' field to domain administrator.

The above is the required step.

VMmark Wkld Setup failed mailserver 'remove old files' and 'do restore failed'

psmith2006 — Mon, 05 May 2008 19:57:41 GMT

These steps enable additional logging of debug messages by the STAXmonitor that are in the VMmark Harness. You'll see more verbose messages in the STAXmonitor window and these messages are also written to the results/Results_<datestamp>/STAX*User*.log file. We use this info to help debug problems. It does not effect the functionality or performance of the test it just records operation tracing in the vmmark harness.

Tile to Client Ratio

psmith2006 — Tue, 18 Mar 2008 18:29:26 GMT

read somewhere in the documentation that a client system could run as the client for more than one tile if host aliases were setupproperly. When looking to do 8 or more tiles I'm starting to wonder if I can do
2-3 tiles per client and if that would give me accurate results?

It would depend heavily on the processing power of the "CLIENT" and the amount of memory and number of NICs it had.

Also.... can loadsim run its .sim files on a different host than the test was initialized under? That seems like the most significant issue technically.

No. they have too much configuration specific data - clientname. mailserver name, domain name, etc.

Default Values in vmmark 1.0

psmith2006 — Tue, 18 Mar 2008 00:28:07 GMT

The vmmarktemplate.sim file to copied to vmmark.sim at the begining of the test by the harness.

These files contain unique encoded info for the particular mailserver they were set up for so vmmarktemplate.sim is used to restore the vmmark.sim file which the harness can use string matching to patch in an updated runtime.

Help! Can't generate tilescore.

psmith2006 — Tue, 05 Feb 2008 21:56:41 GMT

A couple of things:

there was not a database.wrf file in the results.zip

you seem to have SPECweb2005 v1.20 - see http://communities.vmware.com/thread/124866?tstart=0 for more details.

For the webserver problem check that you have set the client registry parameters as specified on page 55 in the benchmark guide.

If you are having problems it is always a good idea to set DEBUGFLAG=1 in the VMMARK.CONFIG file so that the Results_<datestamp> directory will provide as much info as possible for debugging.

VM system clock drift on Webserver VM's

quikah — Thu, 24 Jan 2008 00:18:16 GMT

I have not noticed any significant clock drift in my runs. Are you using the template VMs or have you built your own? If you built your own be sure you include "clock=pmtmr " as a kernel boot option.

Are you noticing clock drift in any other VMs?

Installation doc?

bherndon — Thu, 13 Dec 2007 02:15:25 GMT

You can download the VMmark kit from www.vmware.com/go/vmmark. The kit contains full documentation.

Javaserver running forever

psmith2006 — Mon, 12 Nov 2007 20:46:28 GMT

Yes, follow the directions on page 126 of the benchmarking guide.
But the DEBUGFLAG has to already be enabled so the most info
gets recorded.

Paula

VMmark score

Mizzou_RobMan — Wed, 26 Sep 2007 21:49:07 GMT

The VROOM blog was very beneficial. Thanks for the info and explanation. I think I was being dense.

http://blogs.vmware.com/performance/

VMmark V1.0.0 Released

psmith2006 — Wed, 22 Aug 2007 14:47:53 GMT

VMmark V1.0.0 was released July 23 2007

For details and downloads see:

http://www.vmware.com/products/vmmark/

For info on how VMmark is being used see:

http://blogs.vmware.com/performance/

New Beta date???

psmith2006 — Wed, 22 Aug 2007 14:41:56 GMT

VMmark V1.0.0 has been released, the beta has closed.

For Details See:

http://www.vmware.com/products/vmmark/

what is VMmark

kimono — Fri, 17 Aug 2007 02:54:14 GMT

It's a framework & methodology for benchmarking virtualization platforms against one another. It's also a great learning tool to see the differences introduced by configuration changes (such as No# VMs per LUN configuration, networking) For example we're using it to benchmark every type of host that comes in the door to build up our database. We see some interesting changes of the results based on storage configuration (particularly). If you wish to publish any results publically, there are pretty stringent rules regarding the configuration you used to benchmark.

Theoretically I see no reason why VMMARK could not be used to benchmark different hypervisors.

VMmark 1.0 HTML results - Time Synch problem

rolandt — Wed, 15 Aug 2007 19:35:22 GMT

Yes, I know all clients being on seperate physical systems is one of several requirements for a valid submittable VMmark result. Publication is not my objective with these tests.

My Time Synch solution.
I setup and am using my Prime Client (called "control", which is yes, yet another seperate physical system) as an NTP server. All clients (client0, client1, etc) and all VMs use "control" as their only NTP source. And, of course VMware Tools Time Synch is disabled in all VMs.

vmark/fileserver workload

kimono — Tue, 14 Aug 2007 03:23:50 GMT

What can cause the RC=1 error? I've tried the workaround as suggested, for example the following causes the error to occur on the fileserver:

MailServer/DELAYTIME="250"
JavaServer/DELAYTIME="540"
Standby/DELAYTIME="170"
WebServer/DELAYTIME="5"
Database/DELAYTIME="360"
FileServer/DELAYTIME="180"

and the following causes it on the webserver

MailServer/DELAYTIME="250"
JavaServer/DELAYTIME="540"
Standby/DELAYTIME="170"
WebServer/DELAYTIME="180"
Database/DELAYTIME="360"
FileServer/DELAYTIME="5"

In my case I feel it's just the host is too slow (2 year old single core with slightly overcomitted RAM, non compliant test I know!) because the exact tiles work fine on our over specced 585's.

Windows R2 CLIENT "Sysprep" breaking NTFS permissions on VCLIENT directory

kimono — Thu, 09 Aug 2007 12:57:21 GMT

Thanks for this. I've used NEWSID before and will probably switch to it for the VMMARKING.

What are actual cygwin requirements???

bherndon — Mon, 06 Aug 2007 05:11:52 GMT

I have made a few (admittedly feeble) attempts to pare down the installed packages. I always end up missing things and fall back to a full install due other concerns. I'd be delighted to see anyone with the time and desire help us find the minimal set. As far as your questions:

a) If the download time is an issue, you might try downloading everything to a local machine once and installing everything from a local mount for all of your clients.

b) I never found the disk usage to be a problem since it's hard to buy a system with less than 80GB these days, but I can see that this might be an issue on older hardware.

c) You do in fact need the compilers in case you ever need to recompile the database or fileserver clients due to some version incompatability. In addition, the cygserver must be running to provide IPC facilities. There are also several dll dependencies, etc. that seem to cascade through. I agree that things like xwindows can probably be omitted. I just haven't tested it to be sure.

VMmark public beta is now available!

rolandt — Wed, 18 Jul 2007 16:36:08 GMT

What components of cygwin are really needed for VMmark?

I know it makes it simple to say install everything. But that a) takes a long time to download, b) takes up a lot of extra disk space (up to 2.5GB) on every client server, and c) includes a lot of packages that are obviously not going to be used (like database and xwindows).

Build process suggestions

psmith2006 — Thu, 22 Mar 2007 15:29:12 GMT

The next beta will have a number of usablity improvements.
For example, for the LINUX VMs, we'll be providing pre-configure
applicance VMs that will be available for download. Installation
of some licensed software and some minor configuration will all
that's needed. Stay tuned.

vmmark/javaserver workload

VMwareUserR — Tue, 13 Feb 2007 20:29:09 GMT

The results from debug has sent to vmmark-info.

vmmark/mailserver workload

VMwareUserR — Mon, 12 Feb 2007 22:57:13 GMT

I checked. I can ping both systems from both client and server systems.

vmmark/dbserver workload

psmith2006 — Mon, 12 Feb 2007 19:37:12 GMT

Good.

vmmark configuration setup

psmith2006 — Sun, 11 Feb 2007 23:25:06 GMT

Can you be more precise as to what webserver setup or configuration files are you referring to?
Thanks.

There will be some improvements in the next beta but there are always going to be
some edits needed since everyone's test lab isn't going to be identical.

Configurations - /cpu/ram/disk

jenniferanderson — Wed, 07 Feb 2007 05:28:10 GMT

No, virtual CPUs have no notion of hyperthreading. ESX Server takes advantage of hyperthreading on the box, but the virtual CPUs are not considered hyperthreaded. So you can think of a virtual CPU as being similar to a single-core non-hyperthreaded CPU.

running a subset of the workloads

psmith2006 — Wed, 07 Feb 2007 01:23:37 GMT

The VMmark harness does initiate tbench_srv.exe on the client.
You may want to tweek FileServer/TSRVDELAY in VMMARK.CONFIG
to 10 or 20 and try again. If that doesn't work, then check the
message in the fileserver0.wrf file. If you get a can't bind message,
try rebooting the client and try again. You can also try changing
the FileServer/base_port to some other value tahn 1066.

Let me know if it works or not.

problem with build - apc for webserver

psmith2006 — Mon, 29 Jan 2007 20:37:17 GMT

According to the APC forum there is a fix for this in 3.0.12p2

http://pecl.php.net/package-info.php?package=APC&version=3.0.12p2

See if that helps or fall back to 3.0.11

Standardization?

vmakhija — Tue, 02 Jan 2007 23:03:45 GMT

Doug,

Thanks for your comments. Indeed it would be nice to have pre-configured VMs, but as you point out the licensing restrictions prevent that. For the RHEL VMs we are looking into ways of improving the setup time, but even there the applications (Oracle, SPECweb etc.) would still have to be installed by the end user.

Luckily, the investment in time for the VM creation is a one time cost. We have found that the cost of creating additional tiles is much less compared to the first tile as most of the existing VMs can be cloned and their configuration modified (as opposed to creating everything from scratch).

Additionally, some of our partners have even cloned the clients to further reduce setup time for additional tiles. Having said that though, we are continuing to look into ways to reduce the setup time.

Thanks,

Vikram.