We had an odd issue come up today in our VSphere installation.  We have some HP C7000 chassis with a number of BL465 blades (mostly G7 and G8 models) with 128 or 256GB of RAM.  All are running the HP customized image of VSphere 5.1 update 2.  All drivers and firmware are up to date.  Over the last few months we've been switching our systems over to using 10Gb uplinks into our Cisco Nexus 5000 infrastructure.  We would migrate all VM's off of a chassis, rebuild the chassis, blades, and VMware configuration, do some verification, then migrate the VM's back to the chassis.

Each host has four 10Gb uplinks.  Two are dedicated to the VM networks and are on a virtual switch with the various port groups/VLAN.  Each uplink goes to a different core switch.  Management and VMotion are on a separate vSwitch, each with it's own 10Gb port, with the vmotion and management nic being "standby" for the other.  We're not using distributed switches at this time.

Monday afternoon we got reports that one of the Windows 2008R2 VM's was showing poor network connectivity.  Pinging it seemed to be OK, but trying to connect to it with RDP or other management tools was erratic.  Connecting to the console through VCenter we saw that the Windows logs were showing warnings related to network communications, talking to the domain controllers, etc.  Nothing showed any excessive load on task manager or in VMWare's performance tabs.

After checking a few things (and not finding anything), we tried migrating the VM to another blade.  Immediately everything on the VM started working correctly.  Thinking it was a blade issue, we checked some other VM's on that host.  A couple showed similar issues, but most seemed OK.  

We moved the original VM back to the "problem" blade, and it continued to work OK.  Migrating the other VM's showing issues "fixed" them, even if they were moved back to the original blade.

We checked some VM's on other blades in the same chassis, with the same configuration, built the around the same time, but they all seemed OK.

This chassis and blades have the same configuration we set up on another chassis a few months ago; that one has been working fine.  The other blades in the same chassis seem to be OK.

I've put the blade in maintenance mode until I can figure it out (or just rebuild it), but I can't figure out what would be causing this type of behavior.

Any help would be appreciated.

Hi,

I think in ESXi 4.x era that one can delete a independent persistent disk from a VM that has snapshot , but attempt to do so in esxi5.1 will result in following error:

"Cannot remove virtual disk from the virtual machine because it or one of its parent disks is part of a snapshot of the virtual machine."

This is very annoying as I need to delete the snapshot , delete the disk and redo a snapshot , of which doesn't even contain the said *Independent* disk.

Is this a bug? can someone comment if it can be addressed?

Any ideas why we got following alert -message between 1 hour...  This alert is also displayed in vSpehere Client Events  lists.

I did not find reason for this but I hope that somebody can help me... :-)

This alert "Error(1) saving state to /bootbank" is otherwise similar in every hour but cpu -number changes every hour...

Hi experts

I'm doing an online addition of a logical drive from Vsphere command line;

esxcli.exe -u root -p password --server=10.x.x.x hpssacli cmd -q "ctrl slot=0 create type=ld drives=2I:1:5,2I:1:6,2I:1:7,2I:1:8 raid=1+0"

in below output, Logical Drive :1 is existing drive and 2 is newly created... New drive is not getting detected in ESXi until unless do a Restart...

Is there I'm missing something here or it does not support online addition of logical drives ??

ESXi Version: 5.5.0 (recently upgraded from 4.x)

HP Proliant DL 380 G7

HP Smart Array P410i

C:\Program Files\VMware\VMware vSphere CLI>esxcli.exe -u root -p root --server=10.x.x.x hpssacli cmd -q "controller slot=0 ld all show detail"

Smart Array P410i in Slot 0 (Embedded)

   array A

      Logical Drive: 1

         Size: 273.4 GB

         Fault Tolerance: 1+0

         Heads: 255

         Sectors Per Track: 32

         Cylinders: 65535

         Strip Size: 256 KB

         Full Stripe Size: 512 KB

         Status: OK

         Caching:  Enabled

         Unique Identifier: 600508B1001C8AF8AB99C0301DDD442B

         Disk Name: vmhba0:C0:T0:L0

         Mount Points: None

         Logical Drive Label: A00169285001438009BF78A0CE46

         Mirror Group 0:

            physicaldrive 1I:1:1 (port 1I:box 1:bay 1, SAS, 146 GB, OK)

            physicaldrive 1I:1:2 (port 1I:box 1:bay 2, SAS, 146 GB, OK)

         Mirror Group 1:

            physicaldrive 1I:1:3 (port 1I:box 1:bay 3, SAS, 146 GB, OK)

            physicaldrive 1I:1:4 (port 1I:box 1:bay 4, SAS, 146 GB, OK)

         Drive Type: Data

         LD Acceleration Method: Controller Cache

   array B

      Logical Drive: 2

         Size: 558.7 GB

         Fault Tolerance: 1+0

         Heads: 255

         Sectors Per Track: 32

         Cylinders: 65535

         Strip Size: 256 KB

         Full Stripe Size: 512 KB

         Status: OK

         Caching:  Enabled

         Unique Identifier: 600508B1001CD192F839C181FC21A41C

         Disk Name: unknown

         Mount Points: None

         Logical Drive Label: 064E1DDD5001438009BF78A0AB7D

         Mirror Group 0:

            physicaldrive 2I:1:5 (port 2I:box 1:bay 5, SAS, 300 GB, OK)

            physicaldrive 2I:1:6 (port 2I:box 1:bay 6, SAS, 300 GB, OK)

         Mirror Group 1:

            physicaldrive 2I:1:7 (port 2I:box 1:bay 7, SAS, 300 GB, OK)

            physicaldrive 2I:1:8 (port 2I:box 1:bay 8, SAS, 300 GB, OK)

         Drive Type: Data

         LD Acceleration Method: Controller Cache

Hardware: DL360 G6

RAID Controller: P410i w/512MB cache, No Battery

Drives: 4 SAS Drives (RAID Config Varies for testing)

CPU: 5530 Xeon

RAM: 48GB

So I am testing different RAID configurations and checking the performance.

I started with 2 drives in a mirror (~300GB), then 3 drives in in RAID 5 (~300Gb) and I get the attached error message.

"Parted util failed with message: Error: The primary GPT table states that the backup GPT is located beyond the end of disk.....etc"

Now, I continue to add the 4th disk for a RAID 5 (~900GB) test and it runs. Then I convert to a RAID 10 (~600GB) and I get the same error.

Looks like whenever the useable disk space is ~300GB or ~600GB it bombs out.

Each time I change the RAID type, I clear the configuration then create a new aggregate of disk then a new RAID set. So *theoretically* there should be no residual data on the LUN.

Strangely, when I was able to get it installed and a VM to test disk I/o via SQLIO I always got only between 15-16MB/s... That is not right!

Any ideas to fix this error?

Thanks

Hi there,

i hope to be quite at the correct spot at the forum to find help and be able to describe the problem as detailed as possible.

The Situation:

We're trying to run Windows 8.1 with ESX 5.5, on first spot there seems to be no problem, but after some "virtual working" the magic begins...

Current testing Hard-/Software:

ESX Server:

Intel(R) Xeon(R) CPU E5-2650 v2 @ 2.60GHz

RAM: 262.111 MB

ESX-Version: VMware ESXi, 5.5.0, 1331820

VM Host: (also created in different variations to eliminate configuration errors)

Guest OS: Microsoft Windows 8 (64-Bit)

Compatibility: ESXi 5.5 und höher (VM-Version 10)

RAM: 8 GB

SCSI-Adapter: LSI Logic SAS

HDD: 128 GB (on local RAID 6 over LSI Controller)

Possible Workflow - (magic bug) - This is just one way to generate our Problem:

- Windows 8.1 is up to date

- In the example i use Firefox, but it's not dependable on the software used, IE produces the same phenomen. Downloads or new files created with other software also _CAN_ (<- evil - magic word -) produce this problem.

- Try to download something (strangly, the problem occurs very accidentaly - here i tried to download some default software)

(You may can't see the problem, buwhahaha ...yeah... but this is what i currently see, nothing, exactly... the download hangs up at "1 second")

- The file-writing process stopped

- Trying to access the downloaded - incomplete - file: "explorer.exe" stopped reacting

- Closing Firefox (kill it!) releases the filehandle and i can rename the downloaded ".part".-file -> works!

- The temporary ".exe" file (firefox created) is existing (0 byte). The Process of "moving" the temporary download file ".part" to the finished ".exe" file seems to fail?

- Downloading the same file again can or cannot work, its strangly randomly

- Downloading a file with internet explorer may already stop at any point between 1 to 100% download, firefox always hangs @ 99 or 100%

Now you may think thats not a VM Problem... but...

The same problem does occur working with microsoft office products or other third party tools like yEd.

Also, there are no microsoft log entries except the "Application Hang" that Firefox returns...

We already changed the ESX-configuration to default, used another server (other hardware!).

We reinstalled Windows 8 ... let me recount... countless

We have no problems with other OS like (Windows Server 2012 R2 (i thought that 2012 may have the same problem.. but nope!), Windows XP, Windows Server 2008 R2, Windows 7, Ubuntu, Centos, Red Had, SuSE, ...)

Writing files to a network share does not result in this error!

Also copy local files or install programs either.

Its only the creation of new files by (any) software!

I think this will only happen if the Software write temporary files and moves them afterwards.

Do you have any expirience with this bug or a hint, where to look for log entries / settings?

Except this filehandling-bug Windows 8.1 performs fine

Thank you all in advance, i want to try the best to fix this and let you know the results.

Edit:

Added the possible solution to use another browser or software, this will not help since the bug is appearing with every Applikation that writes files. I hope this will clarify the facts for future discussions

Edit 2:

We currently have an image that has been built using CentOS op sys.  Originally it was set to have two VMDK disks of 80GBs each.

However we have now managed to optimise the CentOS operating system down to 40GBs each, however we are having problems trying to shrink the VMDK files down.

If we use VMWare converter and change the values of the VMDKs and then click ok and convert they are still the same size.

Please if someone could point me in the right direction to resolve this issue that would be great.

I would have thought that there would perhaps be a third party application that would be able restore the CentOS server into a smaller VMDK file.

Any assistance would be greatly appreciated.

Hi all,

Hitting an issue in our environment that may not be resolvable but wanted to reach out to folks and see if they found a way to override.

Currently have vm monitoring enabled in our cluster. Thing is we would look to reboot VM's whose VMware tools are showing a status of "not running" - now the key thing is by most definitions the system is available when this happens, can be pinged, RDP'ed etc but we would still like them rebooted as it ultimately resolves the issue we see (eg unable to connect via citrix to the machines in question)

Duncan in the post below points out in order to avoid false positives any storage/network IO activity will be checked after heatbeats have failed in order to "double check" there is a problem with the VM

VM Monitoring (aka VM HA) heartbeat - Yellow Bricks

What I was wondering was is there a way to override the storage/network IO activity check and thus reboot the vm's as soon as the VMware tools stops running on them?

Many tx for any thoughts

Hallo,

wir haben unser SAP Produktivsystem virtualisiert. Nach der Virtualisierung haben wir schlechtere Schreibwerte beim Festplattenzugriff und Geschwindigkeitsverluste beim Upload von Daten in unser SAP System.

Unsere Umgebung:

3 ESX Hosts - DL360p G8 - 2x Intel Xeon 8 Core E5-2690 2,90 GHz – 132 GB RAM
ESXi 5.1 - 2000251

Storage:

iSCSI Storage Dell EQL PS4100 und PS6100

Kann jemand hierzu etwas sagen ob VMware hier Probleme verursacht?

Oder was die Besten Einstellungen sind um das System optimal am Laufen zu haben!

Danke für Eure Rückmeldungen.

Grüsse

Krischtof

in English - Sorry my English is not so good....;-)

we have virtualized our SAP production system. Once virtualized, we have worse write performance on the disk access and speedloss when uploading data into our SAP system .

Our environment :

3 ESX hosts - DL360p G8 - 2x Intel Xeon 8 Core 2.90 GHz E5-2690 - 132 GB RAM
ESXi 5.1 - 2000251
Storage :
iSCSI Storage Dell EQL PS4100 PS6100 and
Can someone say something to this problems about the performance? Does VMware is making the problems here ?
Or what the best settings are to have the system running optimally !
Thank you for your feedback .
regards

the SAP-VM is a new machine (no P2V) with windows 2008 datacenter, 8 Cores, 32 MB - SAP Database is DB2

Hi,

We've been running a VMware instance with Windows Server 2008 R2 as a terminal server, but the company has grown a lot and there's just not enough cpu/memory to keep it working smoothly.
We have installed a completely new VMware server and I'd like to clone this terminal server to the new VMware server but use both of the terminal servers.

Can someone recommend me a method how to do this?

Thank you,
Dennis

Community,

I could use some help on thoughts, currently there are a few projects about upgrading network/storage components etc. but I was thinking on some different approaches.

Currently our setup of 175 ESXi 5.1 hosts is roughly :

CPU/Memory:

Depending on model, during lifecycle we have about 4 different models active (same HW vendor), each has the NIC/Storage configuration below:

Storage:

2 x 4Gb FibreChannel (Connected to different fabrics)

Networking:

8x 1Gbps NICs

      - 2x vMotion (on different switches)

      - 2x Management (on different switches)

      - 4x VM Traffic (on different switches)

So what to do in the future? Currently networking is mostly bound on 1Gbps because the ToS which aren't upgraded yet.

There aren't any plans about building a converged infrastructure etc. but I was thinking about it.

We don't have any real problems which will push us in one direction. Because the fragmentation of disciplines every team has another point of view how to do things.

Case 1

Keep the current design, but upgrade the Network and storage components when needed in the future. This is the way we work now.

Case 2

Have a look at a converged infrastructure solution? Bladecenter / infiniband ?

Case 3

What about network virtualization (VMware NSX) with a converged solution (or not?)

A few things I want to achieve (no business requirements (yet))

- Increase/Maximize performance

- Low(er) latency

- Decrease costs (cables, DC space, power, cooling, administration) in the future

What are your experiences, troubles, things to think about, what not to do?

I don´t necessarily need a full explanation (would be nice of course) but if I can get some new thoughts or directions from your answers I could dig in it a little more myself. Just wondering depending on experiences if it's worth to put more effort in it.

Don't worry about any investments, if it's worth to build a Business case which can help saving costs, management etc. it might be worth it.

Hope to hear from you guys

vmware ESXi5.1 disk read and write activity is high

Hi,

platform esxi 5.5.

I have server virtualized Red Hat AS 4 but at boot, after ceck "Enabling swap space", it remains locked.

I have enough space in the datastore.

What does it depend?

Thanks.

hello,

is there any opensource VM backup solution available ?

thanks

Hi Everyone,

          I am trying to upgrade  ESXi from 5.0 to ESXi 5.5 through ISO image but after booting its not showing upgrade option in console.Its showing only Install keeping vmfs data store

and Install overwrite VMFS data store. PFA screen shot. Any help on this will be much appreciated.

Thanks

vm2014

Hi,

When ever when i am trying to upgrade ESXi host upgrade its shows error "ESXi – Cannot run upgrade script on host". I tried to follow the below articles but it didnt help.

http://www.running-system.com/cannot-run-upgrade-script-on-host-during-upgrade-from-esxi-5-1-to-esxi-5-5/

  http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2014084

Thanks

vm2014

Hi,

I cloned a VM. Then i tried to power on the  VM. But i get this error - All required agent virtual machines are not currently deployed on host"

I am seeing this error of first time.

Please help.

Hardware:  Dell VRTX, 3 M620 Blades, 2-10 Core CPU's/Blade (40 Logical Cores Total/Blade), 128 GB RAM/Blade (Will be upgraded to 192 GB RAM/Blade in the near future.)

Operating System On Each VM:  Windows Server 2012 R2 64 bit

I currently have five (5) VM's running.  I am new to VM's and I was very generous in CPU/RAM/Disk Space Allocation when I setup these VM's.  I have been told to add three (3) additional VM's.  One (1) of these three (3) new VM's should very light since the installed software won't take up resources.

The software vendor for all three of the new VM's talks about one portion of their installed Apps on one of the VM's sometimes having problems with CPU contention, if more than one (1) CPU is allocated for the VM where the software is installed.  The software Vendor talks about VM's using more resources, caching, etc.  The software vendors shows the CPU Requirements for a Physical Server as:  Intel Pentium Dual Core Processor (2 GHz).  This software vendor also refers to more CPU allocation sometime degrading performance.

My Current CPU & RAM Allocations are:

VM #1:  8 Logical Cores - 64 GB RAM

VM #2:  4 Logical Cores - 16 GB RAM

VM #3:  4 Logical Cores - 32 GB RAM

VM #4:  4 Logical Cores - 16 GB RAM

VM #5:  8 Logical Cores - 32 GB RAM

Proposed Future VM's:

VM #6:  4 Logical Cores - 16 GB RAM

VM #7:  4 Logical Cores - 16 GB RAM

VM #8:  2 Logical Cores -   8 GB RAM

VM #7 is the one with the software that the Vendor is concerned about.

Currently:

Blade #1 has VM's 1 & 2

Blade #2 has VM's 3 & 4

Blade #3 has VM 5

From your experience:

1.  What resources issues have you came across related to CPU contention, caching, RAM usage?

2.  Does what I propose above seem like it would have Resource Contention Problems resulting in the system getting bogged down?

3.  What are the best ways/tools to monitor Resource Contention/Resource Usage?

4.  If I wanted to briefly shut down a VM and decrease the CPU Cores and/or RAM Allocated, does this have any negative impact on Windows Licensing?  It seems like it would have no impact on Licensing since Windows should see only a minor hardware change(s).

Thanks in Advance.

Hello all!

I'm experimenting with ESXi on my home computer, but keep running into the "No Network Adapters" error, which I understand is because the 5.5 default drivers don't support the Intel I218-V on my Asus H97M-Plus motherboard.

I've looked through these forums and others, and understand that I can use the ESXi Customizer v2.7.2 to add the Intel drivers. The latest version I can find in a .VIB file is net-e1000e-2.3.2.x86_64 which don't seem to work. The latest version I can fine on Intel's site is e1000e-3.0.4. I've tried using the Customizer to insert that version without success.

Also, I'm trying to run ESXi from a bootable USB drive which I create from the ISO using Rufus-1.4.9

My system specs are:

Asus H97M-Plus

Intel 4790S

32GB DDR3-1600 RAM

What's the next step?

Thanks,

How do you get AD authentication to work with a 5.1 Vcenter Appliance? I went to the administration and then to the Authentication tab. I successfully enabled AD. However, it's not working. Can't log in and it doesn't see my domain if I try to add a permission.