vdiiomark Tracker

Anyone have issues running RAWC 2.0?

vdiiomark — Tue, 05 Jul 2016 16:58:11 GMT

I am setting up a RAWC 2.0 environment, having previously used RAWC several years ago. The problem I am hitting is a pathing problem with Acrobat 11.0 . This is a 32 bit app, hence installs to "C:\Program Files (x86)\" path, however RAWC continues looking in the 64 bit directory of "C:\Program Files\". There is no 64 bit version of Acrobat 11, hence this path cannot be valid. I have tried creating a symlink, but this did not seem to solve the problem. Anyone else hit this?

Re: Strange vCenter Bug

vdiiomark — Fri, 19 Jul 2013 22:18:55 GMT

Thank you so much for your help, it seems to be an Ubuntu problem. I attempted the same operation with a windows box, and it worked swimmingly. I guess I should go over to the ubuntu forums and talk to them.

Re: Strange vCenter Bug

vdiiomark — Fri, 19 Jul 2013 17:07:18 GMT

I have tried both hot adding and adding when the vm is powered off, both result in the same issues. Only one new disk is recognized inside of linux,as apposed to the ten or so new disks. Adding the disks is changing something, it adds that single disk, but nothing more than that. I have rebooted multiple times and still no change. I did make a mistake adding one disk, causing a file lock, but after that, the hot adds seem fine. Consider line 2390, it seems to add fine. What do you mean by your last question. Thanks for the help so far.

Re: Strange vCenter Bug

vdiiomark — Fri, 19 Jul 2013 16:05:01 GMT

Thank you for your time and consideration, this has really confounded me.

When I say delta disk, I am referring to a call I used in the sdk ReconfigVM_Task() where you can pass an argument that allows for the creation of delta disks. http://pubs.vmware.com/vsphere-50/index.jsp#com.vmware.wssdk.apiref.doc_50/vim.vm.device.VirtualDisk.FlatVer2BackingInfo.html

Delta disks point back to a read only vmdk and record and changes between the original disk and the current state. All the disks that I have added point back to the same original vmdk.

I have attached my logs, let me know if you need more details.

vmware.log - Pastebin.com

Strange vCenter Bug

vdiiomark — Thu, 18 Jul 2013 21:45:15 GMT

I ran across a very strange bug in vCenter/vSphere. I partitioned 10 delta disks that all point back to the same backing. All of the disks insist that they exists in vCenter as seen below.

However when I went to manipulate the disks in Ubuntu, only one of the newly partition disks was showing up!

There only exists /sda* and /sdb* in /dev as well. Does anyone have any clue as to what is going on or how to circumvent it?

Possible Bug / Implementation Fault with Adding Delta Backed Disks

vdiiomark — Tue, 16 Jul 2013 21:06:57 GMT

Through the perl SDK I have created a number of delta disks on a virtual machine that all point back to the original backing, not through each other. The disks seem to work wonderfully, however when I go to delete them I get a handful of faults. It seems that the first delta disk that is deleted also removes the primary backing, and the other disks have no where else to point to, and will throw faults when you try and remove them, citing a lack of the primary backing. Even if I attempt a batch operation, deleting them all at once, a fault will be thrown saying I can not complete the operation that way as well. Is there any way to delete the delta vmdks with out deleting the primary backing? Alternatively is there a way to not have the primary backing be deleted until the last delta disk backed on it is removed?

Re: Did I find an API bug? : Adding multiple disks soap error

vdiiomark — Thu, 13 Jun 2013 20:41:09 GMT

I did a dump like you said and it was very helpful. I needed to add the code below to make it work

Turns out that I need a unique key for every disk that I am adding. Thank you for your help

my $disk = VirtualDisk->new(

controllerKey => $devSpec->device->controllerKey,

unitNumber => $devSpec->device->unitNumber,

key => -1 *($i + 1),

backing => $devSpec->device->backing,

capacityInKB => $devSpec->device->capacityInKB,

);

$devSpec = VirtualDeviceConfigSpec->new(

operation => $devSpec->operation,

fileOperation => $devSpec->fileOperation,

device => $disk,

);

Is there a better way to do it than this

Re: Did I find an API bug? : Adding multiple disks soap error

vdiiomark — Thu, 13 Jun 2013 19:19:07 GMT

I dont think so, If that was true, why could I add the three individual machines by array index as apposed to passing it the array of three items

Re: Did I find an API bug? : Adding multiple disks soap error

vdiiomark — Thu, 13 Jun 2013 18:36:42 GMT

I did not want to do a code dump, but here you go

package VManagement;
# Paul Givens
# Evaluator Group
# 6/11/2013
# A package for manipulaing virtual machines through vSphere
# Public functions
# addDisks
use strict;
use warnings;
# I do not know why this lib is needed, delete it for a fun time
use lib "/usr/lib/vmware-vcli/apps";
use Exporter;
use VMware::VIRuntime;
use AppUtil::VMUtil;
# Making a module
our @ISA = qw( Exporter );
our @EXPORT_OK = qw( addDisks );
our @EXPORT = qw( addDisks );
# Declarations
sub addDisks;
sub atomicAddDisks;
# addDisks
# a function to add disks to a virtual machine
sub addDisks{
# Getting arguments
(my $viServer, my $login, my $password, my $verbose) = @_;
# Setting the global variables
$ENV{VI_SERVER} ="$viServer";
$ENV{VI_USERNAME} ="$login";
$ENV{VI_PASSWORD} ="$password";
# Connecting
Opts::parse();
Opts::validate();
Util::connect();
# TEST start
my @diskSizes = (1,2,1);
my @datastores = ("zebi1n", "zebi1n", "zebi1n");
# TEST stop
if ($verbose) {print "Connected to server\n";}
atomicAddDisks("paul-vm1", \@diskSizes, \@datastores, $verbose);
# Disconnecting
Util::disconnect();
}
sub atomicAddDisks{
# Getting arguments
(my $vmName, my $diskSizesRef, my $datastoresRef, my $verbose) = @_;
# Dereferencing
my @diskSizes = @$diskSizesRef;
my @datastores = @$datastoresRef;
# Kb to Gb conversion factor
my $scsiNum = 0;
my @devSpecs = ();
my $vmView = Vim::find_entity_view(view_type => 'VirtualMachine',
filter => { 'name' => $vmName } );
my $vmDevices = $vmView->config->hardware->device;
my $diskFilename;
my $diskFilenameIncrement = 0;
my $diskFilenameCur;
my $diskNum = 0;
foreach my $vmDevice (@$vmDevices) {
# Adjusting the diskNum for disks already present
if (ref($vmDevice) eq 'VirtualDisk') {
$diskNum++;
}
}
# Search virtual SCSI controller
my $controller;
my $numController = 0;
foreach my $vmDevice (@$vmDevices) {
if (ref($vmDevice) =~/VirtualBusLogicController|VirtualLsiLogicController|VirtualLsiLogicSASController|ParaVirtualSCSIController/) {
Util::trace(1, "SCSI controller found : $&\n");
$numController++;
$controller = $vmDevice;
}
}
my $controllerKey = $controller->key;
# Set new unit number (7 cannot be used, and limit is 15)
my $unitNum;
my $vdiskNumber = $#{$controller->device} + 1;
if ($vdiskNumber < 7) {
$unitNum = $vdiskNumber;
}
elsif ($vdiskNumber == 15) {
die "ERR: one SCSI controller cannot have more than 15 virtual disks\n";
}
else {
$unitNum = $vdiskNumber + 1;
}
for(my $i = 0; $i < @datastores; $i++) {
# Conversion to GB
my $disksize = $diskSizes[$i] * 1048576;
my $datastore = $datastores[$i];
# Increase the disk number to represent the machine to be added
$diskNum++;
$diskFilename = "[$datastore] paul-vm1/paul-vm1_" . "$scsiNum" . "_" . $unitNum . ".vmdk";
# Build virtual spec data object
my $diskMode = VMUtils::get_diskmode(
nopersist => "",
independent => ""
);
if ($verbose){
print "SPECS\n";
print "\tVM Name:\t$vmName\n";
print "\tDisk Name:\t$diskFilename\n";
print "\tController Key:\t$controllerKey\n";
print "\tUnit Number\t$unitNum\n";
print "\tDisk Size\t$disksize\n";
}
my $devSpec = VMUtils::get_vdisk_spec(
vm => $vmView,
backingtype => 'regular',
diskMode => $diskMode,
fileName => $diskFilename,
controllerKey => $controllerKey,
unitNumber => $unitNum,
size => $disksize);
###########################################
# Run ReconfigVM method (in VMUtils package)
# with previously defined specs
###########################################
push(@devSpecs, $devSpec);
#increase the unit number for each device
$unitNum++;
}
my $vmSpec = VirtualMachineConfigSpec->new(deviceChange => \@devSpecs);
$vmView->ReconfigVM(spec => $vmSpec);
#my $vmSpec = VirtualMachineConfigSpec->new(deviceChange => [$devSpecs[0]]);
#$vmView->ReconfigVM(spec => $vmSpec);
#my $vmSpec = VirtualMachineConfigSpec->new(deviceChange => [$devSpecs[1]]);
#$vmView->ReconfigVM(spec => $vmSpec);
#my $vmSpec = VirtualMachineConfigSpec->new(deviceChange => [$devSpecs[2]]);
#$vmView->ReconfigVM(spec => $vmSpec);
}

Did I find an API bug? : Adding multiple disks soap error

vdiiomark — Thu, 13 Jun 2013 18:12:22 GMT

I am able to add the disks linearly like below:

my $vmSpec = VirtualMachineConfigSpec->new(deviceChange => [$devSpecs[0]]);

$vmView->ReconfigVM(spec => $vmSpec);

my $vmSpec = VirtualMachineConfigSpec->new(deviceChange => [$devSpecs[1]]);

$vmView->ReconfigVM(spec => $vmSpec);

my $vmSpec = VirtualMachineConfigSpec->new(deviceChange => [$devSpecs[2]]);

$vmView->ReconfigVM(spec => $vmSpec);

However, when I attempt to add all the disks at once I get a SOAP fault

my $vmSpec = VirtualMachineConfigSpec->new(deviceChange => \@devSpecs);

$vmView->ReconfigVM(spec => $vmSpec);

The Error:

SOAP Fault:

-----------

Fault string: Cannot complete the operation because the file or folder /vmfs/volumes/ba5e81fe-201a6c4e/paul-vm1/paul-vm1_0_3.vmdk already exists

Fault detail: FileAlreadyExistsPAULWORK>clear

Is there anyone out there that knows what is going on?

Soap fault on vmdk creation

vdiiomark — Wed, 12 Jun 2013 18:06:50 GMT

I want to create a vmdk on my datastore so I attempted this

sub atomicAddDisks{
# Getting the views
print "DEBUG : GETTING CONTENT\n";
my $content = Vim::get_service_content();
my $VDiskManager = Vim::get_view(mo_ref => $content->virtualDiskManager);
my $diskSpec = VirtualDiskSpec->new(
adapterType => "ide",
diskType => "thin"
);
print "DEBUG : CREATING VDISK\n";
$VDiskManager->CreateVirtualDisk_Task(name => "[zebi1n]helloworld.vmdk", spec => $diskSpec);
print "DEBUG : AFTER VDISK\n";

As you can see, I hardcoded the path for debugging but I get a very unhelpful SOAP error

SOAP Fault:

-----------

Fault string:

Fault detail: InvalidArgumentFault

Does anyone know what is going on?

Trying to create a VMFS datastore

vdiiomark — Tue, 28 May 2013 22:48:17 GMT

I can not figure out what I am doing wrong, a very unhelpful error "Undefined subroutine &VmfsDatastoreCreateSpec::type called at /usr/share/perl/5.14/VMware/VICommon.pm line 1501" is thrown when I run this subroutine

sub attach(){

#getting the arguments

(my $storage, my $host) = @_;

#creating the arguments for creating the datastore

my $partition = HostDiskPartitionSpec->new(

chs => undef,

partition => undef,

totalSectors => undef

);

my $hostPartition = HostScsiDiskPartition->new(

diskName => $storage->uuid,

partition => 0

);

my $hostVmfs = HostVmfsSpec->new(

blockSizeMb => undef,

extent => $hostPartition,

majorVersion => 5,

volumeName => $storage->uuid

);

my $spec = VmfsDatastoreCreateSpec->new(

extent => undef,

partition => $partition,

vmfs => $hostVmfs,

diskUuid => $storage->uuid

);

#getting the data store manager

my $dataStoreManager = Vim::get_view(mo_ref => $host->configManager->datastoreSystem);

$dataStoreManager -> CreateVmfsDatastore(Vim::get_view(mo_ref => $spec, view_type => "VmfsDatastoreCreateSpec"));

}

Re: Finding the extents of a datastore?

vdiiomark — Tue, 28 May 2013 22:11:36 GMT

I solved this one myself, DatastoreInfo is extended by VmfsDatastoreInfo, which allows you to access HostVmfsVolume which has extents as a field!

Finding the extents of a datastore?

vdiiomark — Tue, 28 May 2013 21:04:58 GMT

I am trying to extract the name of the extents that each datastore has. When I use the Vsphere GUI Client, I can access it through clicking on a datastore and going to the configuration tab, so I know that information exists somewhere. I have spent quite a long time in the API reference, and have not found a way to access the information. Does anyone have any ideas on how I can nab this data?

Re: VMmark 2.5 Mailserver error (loadgen)

vdiiomark — Tue, 26 Mar 2013 23:56:16 GMT

Thanks for all the great input. Last question:

In the "VMmark 2.5 Benchmarking" document, on page 100, under the section entitled "Configure LoadGen", there are 10 steps listed.

Step #10 is "Start the Initilization phase".

SO, the question is, if I already performed this step here during setup, shouldn't initilization be skipped when running the Load Gen independently? There are several options to run LoadGen, one is to initilize only (as above), another is to initialize and run the workload (as you suggested above), and the last option is to run the workload only. It seems to me that the correct option would be to "Run the Workload only", but perhaps I am wrong.

Thanks.

Re: VMmark 2.5 Mailserver error (loadgen)

vdiiomark — Tue, 26 Mar 2013 16:12:15 GMT

I have migrated my vm to another system and tried different storage devices as datastores, in all cases, I get a failure several minutes in.

The only potential bottleneck appears to be CPU related, not disk related. When configured using 4 vCPU's as directed, the CPU load average hovers somewhere around 75%. That is the only variable that seems high, with memory and disk usage appearing fairly normal.

Are there any documented guidelines for recreating the workload using the Load Generator directly?

I believe the parameters are as follows:

User Count = 1,000

Tasks / User / Day = 278

Length of 'Simulation Day' = 2h, 30m, 0s

Scheduled Run Length = 3h, 30m, 0s

Can anyone confirm if these are the correct Load Gen settings to replicate the run created by VMmark? I would like to run an accurate mail load directly, to eliminate other variables.

Re: Question on Primeclient and client

vdiiomark — Mon, 25 Mar 2013 19:35:39 GMT

Great, thank you for your help. I was able to get everything "mainly" working with 1 client serving both roles for a single tile. By that I mean everything except the Email workload. I am fixing this setup issue and will see if it fixes my other outstanding issue. Thanks again.

Re: VMmark 2.5 Mailserver error (loadgen)

vdiiomark — Mon, 25 Mar 2013 19:02:18 GMT

Thank you for the quick reply.

I just migrated my Mailserver vm onto another datastore. Previously I was on an iSCSI attached, hybrid storage system with SSD caching.

I just re-ran the test with similar results. With the run I just performed, I am on a local HBA RAID controller, with read and write caching in front of 4, 15k rpm disks. Minimal but enough for a single workload.

Just to confirm that my latencies are minimal, I am attaching the vCenter perormance log of the results during the run.

Mailserver setup was running from 12:12 until 12:16. Loadgen started at approximately 12:18 and ran about 10 minutes before dying again.

These traces show the following "Write Rate, Read Latency, Write Latency" . I have attached a png capture of these values.

SO, again what I am seeing does not correlate to what the error reported. With both read and write latencies always below 5 ms, the failure makes no sense to me.

VMmark 2.5 Mailserver error (loadgen)

vdiiomark — Mon, 25 Mar 2013 18:02:51 GMT

First, let me state I have read through the forum's. I have all parts of VMmark 2.5 working correctly, except for the Mailserver workload. It does start up, but then exits with an error. In reading other questions, it seems related to answers given about the mailserver responding too slowly.

However, here is the issue. I have no other workloads running (as configured in the VMMARK2.CONFIG file). I am also monitoring my storage via esxtop to see the reported latencies and transfer. During the first 5 minutes (when loadgen is running), the highest read or write latency found was less than 5ms.

The DAVG and GAVG are showing nearly identical results in esxtop, all less than 5ms along with the QAVG/cmd column showing 0.01 or less at all times. Historical info for the datastore in vCenter shows similar results. So, by all accounts nearly no storage queueing or latency. With no other jobs running, the CPU and memory on the system are also very low.

So my question is why does the Mailserver workload fail? Any pointers much appreciated. Attached is a zip file of my "Mailserver" folder output from the clients "c:\vclient\mailserver" directory.

Question on Primeclient and client

vdiiomark — Sat, 23 Mar 2013 16:24:15 GMT

I am attempting to get VMmark v2.5 running. To start with, I only want to run 1 tile.

I have read every single page of the vmmark 2.5 documentation many, many times. A question I have, that I do not see answered anyplace is this:

"How many total clients do I need?" That is, how many "regular + prime" clients are needed? Is it 1, or is it 2? Can the primeclient act as a regular client for a single tile run, or do I still need another client?

Answers, suggestions and pointers to answers in the documentation I may have missed are greatly appreciated.