I've come across documentation and discussions that state you should always have a certain percentage of free space for all datastores containing vm data. Well, what if that datastore only holds a single data thick provisioned vDisk that I am not snapshotting? The datastore will be used to hold the actual database for a SQL server. It's a 500GB LUN and I want to max out that space when I create this datastore and then the 500GB thick provisioned vDisk. What should I be aware of?
Thanks
KJ
Be aware there is some space need to hold information about the VMFS datastore so you will never be able to use the entire datastore but the is nothing that would prevent you from creating the virtual disk in the remaining space as long as it is thick provisioned and you are not planning any snapshots -
Don't forget vm swap file and vm log files. That's another space you need on datastore. Vm swap file equal vm RAM size - memory reservation.
That can be avoided by creating a seperate virtual disk on a second datastore 0
The device "magically appeared" today and I'm able to create the VMFS store. Thank you both for the insight. Maybe vSphere was a little too busy when I tried to create it?
Its always nice when problems fix themselves 
Glad its working!
Hi Avinash,
thanks for your time and tests ! If you want more information or have me do some more testing, please feel free to ask. Have you had the time to check with our io limit problem ?
Thanks,
Vincent.
I have sever Dell R720 Running ESXi 5.1 these servers use FCOE on intel 2P-x520 10 GB modules
And cisco Nexus 5500 Switches for FCOE and networking.
When I upgraded to 5.5 from 5.1 the WWNs changed from the physical ones to the virtual wwns as shown on the attached screen shot.
Both the 5.1 and 5.5 install used dells OEM of the ESXi installation. I used update manager to upgrade the server.
My questions are:
What are the virtual WWNs used for?
Why would the install change from the physical to the virtual ones?
IF I rezone to the virtual WWNS what problems will that cause?
I have cases open with both VMware and dell but I am not getting very far. I was hoping someone else would be able to shed some light on the issue
I have sever Dell R720 Running ESXi 5.1 these servers use FCOE on intel 2P-x520 10 GB modules
I would recommend to use SDRS with the specific VM settings to manual.
Venkat
I don't want SDRS to move that VM or to have another VM moved to this datastore.
Remove the datastore from SDRS
Are there issues if a single datastore lives outside of the datastore cluster?
No Issues. You won't get the benefits of SDRS.
Hi,
What is your RAID configuration on the P2000? Is your datastore spread across all disks? Ideally you'd like to have a datastore across all of your spindles.
One idea I can think of is for you to Storage vMotion one of VMs suffering the latency to the local disks of the ESXi servers to see if it helps.
Eyal Traitel
Director of Marketing
Reduxio Systems, http://www.reduxio.com
c: +972-58-7300-749
yes the log are almost the same
thank you
2014-01-13T11:53:40.228Z cpu18:33974)Partition: 423: Failed read for "naa.600c0ff0001978ecbbe9ca5101000000": I/O error
2014-01-13T11:53:40.228Z cpu18:33974)Partition: 1003: Failed to read protective mbr on "naa.600c0ff0001978ecbbe9ca5101000000" : I/O error
2014-01-13T11:53:40.228Z cpu18:33974)WARNING: Partition: 1112: Partition table read from device naa.600c0ff0001978ecbbe9ca5101000000 failed: I/O error
2014-01-13T11:53:19.953Z cpu50:33499)ScsiDeviceIO: 7402: Get VPD 80 Inquiry for device "mpx.vmhba0:C0:T0:L0" from Plugin "NMP" failed. Not supported
2014-01-13T11:53:19.953Z cpu50:33499)ScsiDeviceIO: 7412: Get VPD 83 Inquiry for device "mpx.vmhba0:C0:T0:L0" from Plugin "NMP" failed. Not supported
2014-01-13T11:53:19.953Z cpu50:33499)ScsiDeviceIO: 7448: Get VPD 86 Inquiry for device "mpx.vmhba0:C0:T0:L0" from Plugin "NMP" failed. Not supported
2014-01-13T11:53:19.953Z cpu48:33499)ScsiDeviceIO: 6162: Could not detect setting of QErr for device mpx.vmhba0:C0:T0:L0. Error Medium not found.
2014-01-13T11:53:51.503Z cpu54:33980)Partition: 911: Read of GPT header (hdrlba = 1) failed on "naa.600c0ff0001978ecbbe9ca5101000000" : I/O error
2014-01-13T11:54:01.604Z cpu50:33980)Partition: 911: Read of GPT header (hdrlba = 1) failed on "naa.600c0ff0001978ecbbe9ca5101000000" : I/O error
2014-01-13T11:58:11.073Z cpu16:34149)Partition: 911: Read of GPT header (hdrlba = 1) failed on "naa.600c0ff0001978ecbbe9ca5101000000" : I/O error
Hi ,
Im a test engineer in Qlogic India.I have been Testing Qlogic Fibre Channel & CNA VMware drivers for Qlogic and other Vendors for past 3 years.
I have developed interest in VMware OS features like Clusters , NPIVs, Software iSCSI.
I do also have hands on knowledge on Orchestrator Automation.
I m willing to go for a VMware Certification Course to strengthen and enhance my technical knowledge and career.
As I have been primarily associated with storage technologies (also have worked in VMware networking features for Qlogic) I thought this forum might help me to decide the right certification course to take.
Thanks,
Rajsekhar
Hi
I am currently involved in a project in patching our esxi5 hosts one host at a time. I just recently joined a new company so am newer to this organisations environment.
We have 3 seperate clusters all managed in our production vcenter.
I am using VUM to patch a host one at a time using Vmotion to migrate VMs to the other hosts. I can only perform upgrades up to 5.0 Update 3 due to the age of the hosts.
I am aware that we have a few VMs using Physical RDMs, I think i know the answer to this and think I shouldnt expect issue but with my RDM machines which physical compatibilty mode am I right in thinking I shouldnt get a problem migrating them VMs to another host in the cluster? When I go through the stages of peforming a vmotion (but not actually performing a vmotion) when I change host and migrate to another host it does say 'validation succeeded" so in my eyes that means it should migrate over and the RDM disks will been seen on the hosts.
If I do see this message am I right in thinking the machine should have no issues once moved to another host?
Thanks
You're right. In general, if your storage zoning configuration and presentation of your block layer storage devices (aka LUNs, including RDMs and VMFS volumes) is consistent among hosts in your clusters, and thus the target host has access to them, there won't be any issue with vMotioning VMs with physical or virtual mode RDMs and this is supported.
The only exception to this is if you're running in-guest-clusters with shared storage (like MSCS) on these VMs or other applications that use persistent SCSI reservations.
vMotioning of clustered VMs with shared storage (which are using physical mode RDMs) is not supported by VMware:
http://kb.vmware.com/kb/1037959
You can read more on the technical background about this here:
But if you're not using clustering or something similar with the RDMs on the VMs, you're good to go.
Hi Rajsekhar,
The certifications from VMware are primarily towards infrastructure administration and design. However I know some QA eng who prefer to do VCP certification to demonstrate their VMware skills.
Hope this helps.
-f10
With your experience I would suggest working towards a VCP in the Data Center Virtualization track - you can learn more here VMware Certification
Hi All,
I have a requirement where in I have to disable and then enable ATS on a vmfs5 datastore with active I/O. Since it has to happen on a per lun basis, I have ruled out the possibility of using advanced settings /VMFS3/HardwareAcceleratedLocking
#esxcli system settings advanced set --int-value 0 --option /VMFS3/HardwareAcceleratedLocking
I know below command will help me to set ATS on a per lun basis but it throws out error when I use it on a vmfs5 partition with active I/O. Is there any way to get this done.
On a VMware ESXi 5.0.0 build-1311175, I get below error.
~ # vmkfstools --configATSOnly 1 /vmfs/devices/disks/naa.600a098038425863515d426b51596551:1 --force
Checking if all devices backing VMFS '/vmfs/volumes/miio1' are capable of ATS. This may take a few seconds...
VMFS '/vmfs/volumes/miio1' cannot be upgraded to ATS capability as all devices backing it are not ATS capable
Error: Operation not supported
On a VMware ESXi 5.1.0 build-1065491, I get below error.
~ # vmkfstools --configATSOnly 1 /vmfs/devices/disks/naa.600a098038425863515d426b51596551:1 --force
Checking if all devices backing VMFS '/vmfs/volumes/miio1' are capable of ATS. This may take a few seconds...
Failed to check ATS capability of VMFS '/vmfs/volumes/miio1'Error: Read-only file system.
Thanks
A Williams
Hi,
AFAIK this can be done only when the VMs are powered off in the LUN. Not sure if I wrote about this here
http://highoncloud.blogspot.in/2013/12/read-all-about-vaai-overview-setup.html
-f10
Apologies, you probably wont find that information in my blog. However the following KB does state that the LUN should be offline when disabling ATS if the volume is ATS-only, please check this KB for more information VMware KB: Disabling the VAAI functionality in ESXi/ESX
Hope this helps !
-f10