I'm working... i'm not working... i'm working... i'm not working

[thaifusion]

Hello community,

I made it through setting up a whitebox, had the right NIC card sent from the USA to Thailand and did all my homework. I got through it all pretty easy after a few month of due diligence.

I made a rookie mistake by allocating a hair to much memory to each VM, got past that. I'm running, but limping.

Here are my specs:

Whitebox

Gigabyte EP45-DS3R

X3360

8GB of Ram

4 500GB Sata

1 1TB Sata

Install goes smooth, convert of VMs from Fusion goes smooth. allocating storage, etc. All straight forward.

"The thing is that this exact machine was a hackintosh before running OSX server with Fusion and these very VMs and it ran smooth for the last 3 months. Thinking I would get better performance, i migrated to ESXi4. I am not getting better performance and I want to know what is ruining my day."

As you can see in my screen shot, the VMs seem to freeze up on a regular basis.

The VMs each have 1vCPU

VMa - 2GB

VMb - 2GB

VMc - 1GB

I updated the VMWARE tools on each machine

The VMs run pretty close to idle most of the time.

Any ideas? I could really use some help... NIC? MOBO? Settings?

Thank you in advance

[thaifusion]

I want to add that while the machine is HUNG up... if i try to access it via the Console, it is frozen.

[thaifusion]

I'm getting a lot of errors in my logs that look like this (would that freeze up the VMs?):

• The host name used for the connection does not match the subject name on the host certificate

Oct 07 15:52:31.123: player|

Oct 07 15:52:31.123: player| * The host certificate chain is not complete.

Oct 07 15:52:31.123: player| SSLVerifyIsEnabled: failed to read registry value. Assuming verification is disabled. LastError = 0

Oct 07 15:52:31.123: player| SSLVerifyCertAgainstSystemStore: Certificate verification is disabled, so connection will proceed despite the error

Oct 07 15:52:32.014: player| vmClientCore::VMControl::UpdateConnectionState: connection state changed: old state is 1 and new state is 2

Oct 07 15:52:32.014: player| vmClientCore::VMControl::UpdateConnectionState: connection state 2 fired

Oct 07 17:05:55.645: player| SSLVerifyCertAgainstSystemStore: Subject mismatch: localhost.localdomain vs 192.168.1.20

Oct 07 17:05:55.645: player| SSLVerifyCertAgainstSystemStore: The remote host certificate has these problems:

Oct 07 17:05:55.645: player|

Oct 07 17:05:55.645: player| * The host name used for the connection does not match the subject name on the host certificate

Oct 07 17:05:55.645: player|

Oct 07 17:05:55.645: player| * The host certificate chain is not complete.

Oct 07 17:05:55.645: player| SSLVerifyIsEnabled: failed to read registry value. Assuming verification is disabled. LastError = 0

Oct 07 17:05:55.645: player| SSLVerifyCertAgainstSystemStore: Certificate verification is disabled, so connection will proceed despite the error

Oct 07 17:05:56.348: player| CnxConnectAuthd: Returning false because CnxAuthdProtoConnect failed

Oct 07 17:05:56.364: player| Cnx_Connect: Returning false because CnxConnectAuthd failed

Oct 07 17:05:56.364: player| Cnx_Connect: Error message: Error connecting to /bin/vmx process.

Oct 07 17:05:56.442: player| cui::MKS::OnReacquireDone: Failed to reconnect the MKS: Error connecting to /bin/vmx process.

Oct 07 17:05:56.442: player| cui::MKS::ReconnectionFailure: Re-connection failed: Error connecting to /bin/vmx process.

Oct 07 17:05:56.442: player| cui::MKS::ReconnectionFailure: Retrying...

Oct 07 17:05:56.958: player| cui::MKS::OnReacquireAbort: Failed to acquire the MKS ticket: Queued MKS ticket acquisition timed out after 500 ms

Oct 07 17:05:56.958: player| cui::MKS::ReconnectionFailure: Re-connection failed: Queued MKS ticket acquisition timed out after 500 ms

Oct 07 17:05:56.958: player| cui::MKS::ReconnectionFailure: Retrying...

Oct 07 17:05:57.458: player| cui::MKS::OnReacquireAbort: Failed to acquire the MKS ticket: Queued MKS ticket acquisition timed out after 500 ms

Oct 07 17:05:57.458: player| cui::MKS::ReconnectionFailure: Re-connection failed: Queued MKS ticket acquisition timed out after 500 ms

Oct 07 17:05:57.458: player| cui::MKS::ReconnectionFailure: Retrying...

Oct 07 17:05:57.692: player| vmClientCore::VMControl::OnConsolePoweredOff

Oct 07 17:05:57.692: player| vmClientCore::VMControl::DoDisconnect.

Oct 07 17:05:57.692: player| vmClientCore::VMControl::UpdateConnectionState: connection state changed: old state is 2 and new state is 1

Oct 07 17:05:57.692: player| vmClientCore::VMControl::UpdateConnectionState: connection state 1 fired

Oct 07 17:05:57.739: player| vmClientCore::VMControl::UpdateConnectionState: connection state changed: old state is 1 and new state is 1

Oct 07 17:05:57.739: player| vmClientCore::VMControl::UpdateConnectionState: connection state changed: old state is 1 and new state is 1

Oct 07 17:05:59.786: player| vmClientCore::VMControl::UpdateConnectionState: connection state changed: old state is 1 and new state is 1

Oct 07 17:06:00.004: player| vmClientCore::VMControl::UpdateConnectionState: connection state changed: old state is 1 and new state is 0

Oct 07 17:06:00.004: player| vmClientCore::VMControl::UpdateConnectionState: connection state 0 fired

Oct 07 17:06:08.926: player| vmClientCore::VMControl::Connect

Oct 07 17:06:08.926: player| vmClientCore::VMControl::Connect: hostname = 192.168.1.20:443, cloneticket = 5234de50-c53b-13bc-99ca-a9f6505c4ea6, moid = 48

Oct 07 17:06:08.942: player| Resolving IP address for hostname 192.168.1.20

[Dave_Mishchenko]

if you look at the performance logs for a VM do you also see a freeze in CPU or disk activity? Can you post the vmware.log file for one of the VMs?




Dave

VMware Communities User Moderator

New book in town - vSphere Quick Start Guide -http://www.yellow-bricks.com/2009/08/12/new-book-in-town-vsphere-quick-start-guide/.

Do you have a system or PCI card working with VMDirectPath? Submit your specs to the Unofficial VMDirectPath HCL - http://www.vm-help.com/forum/viewforum.php?f=21.

[thaifusion]

HI Dave,

Thank you for looking into this. I have switched motherboards, now the memory. The 3 VMs configured the same way ran so smoothly under fusion with a mac osx server host. I can't imagine why they would run so poorly with a ESXi host... i made the move thinking it was a better host. Now i'm in a bad situation at my office.

Ill get the logs and post them... thank you

[thaifusion]

Dave,

I don't know how to get you the logs, but here are screen shots of the DISK/CPU and Pings during a hang. Of all 3 VMs that were running on this box when it was OSX Server/Fusion Host.. I only have 2 running now and its still hanging.

I'm starting to believe that either something is wrong with my CPU, something is wrong with my install of ESXi, I need a second NIC, or ESXi just doesn't like my hardware and I need to setup a different host or go back to OSX.

Thank you

[thaifusion]

Dave,

I think the bottom line here is that ESXi does not like my hardware. I can't even run a single machine. I'm going to change out the CPU. If that does not work. I'm going to reinstall one time and give it a whirl..but for the most part I think my answer is installing Linux and setting up ESX Server on it (which I was going to do before, but found the web interface very buggy). If that does not work, I will go back to OSX/Fusion. It makes no sense for the machine to freeze up like it does.

[jfelinski]

Please check memory usage... maybe your VM's are swapping with host?

In performance tab add counters for Memory swap in, memory swap out and Memory baloon and let us know.

Regards

---

MCSAS,CompTIA Security, VCP

[J1mbo]

Also try running a clean install of any OS (only) to confirm hardware issue. Were there any drivers like vmware tools installed on the VMs for Fusion? - if so these should be removed and the vmware tools installed.

Please award points to any useful answer.

[thaifusion]

Go Yanks... It's an excellent idea. I'm trying to figure out how to do that now.

thank you...

[thaifusion]

I had removed the old tools and installed new ones from esxi. thank you tho... i will create a fresh machine as well.

[Rumple]

Once you get things running your performance is probably not going to be great anyhow because you probably do not have a SATA controller with a battery backup unit on that.

In write through mode (cache disabled) the performance is functional at best.

I am running on SATA drives plugged into the IDE controller on a server with no cache and while I have some VM's running, but Its sometimes excruciatingly slow...can't always get the console to come up, just trying to manage the VM drives me crazy sometimes.

With something like a megaraid controller with a BBWC, I'd have no problems whatsoever.

[thaifusion]

Along with some of the other insightful comments, that makes a lot of sense. Altho, its frustrating to hear about others who have "whiteboxes" up and running with the same hardware I have with no problems. I know that vSphere is designed to run on more robust hardware and while it runs on these desktop powerhouses, they are light weights compared to the beasts from Dell, HP, etc. I am going to give Linux/ESX Server a shot right now. I'm really bummed, I loved the way vSphere looked, acted, etc. Maybe after my company is profitable (all this work is for my startup), I will get a Dell T710 embedded 😃 drool

[Rumple]

One thing that really makes all the difference is if your motherboard has PCIe slots (like for video cards), get an HP E200 series card with the BBWC on it. They are typically $400-500, but you can hook upwards of 8 SATA drives to it using the SAS->SATA cables...

That will give you more then enough performance...

Unfortunately more me my server only has PCI-X slots so I am SOL...

[DSTAVERT]

Servers make terrible laptops and aren't that much better as desktops. Servers are put together as a system who's primary job is to serve. It is more about all the pieces of the package working towards that goal. I wouldn't want to have one of those high performance video cards in my server.

You can find deals on real servers that aren't much beyond what a whitebox costs to build. New last years models, factory returns and refurbished machines are usually offered by manufacturers ad very favorable prices. You won't regret your first real server. Inexpensive even in the short term.

[thaifusion]

Hi,

Working with VMWARE has been a great experience. As my use grows, I will need the proper hardware to achieve the results I am looking for. I had some extra hardware and after doing some reading, felt comfortable that I could build a whitebox as I migrate from Fusion to more of an infrastructure. I have come to realize that vSphere must be run on proper hardware and a whitebox is NOT for production and should only be used to play with.

As a solution, until the business can justify the expense, I put Debian on two boxes using the hardware discussed in this thread, VMWARE Server 2, and everything is running very smoothly. I think this solution will be good for the next year and then I can look at getting something like a Dell T710.

Look forward to rejoining the community at that point and thank you guys for all the help

Cheers!!