Reformat to VMFS 5 LUN

Similar Questions

• Virtual machines with VMFS and LUN VMFS-no?

  Is it possible to attend the ESX3.5 VMFS and Non-VMFS LUNs virtual machines simultaneously?  We are totally an environment totally VMFS/Fiber Channel VI with Windows o/s.   Although we have all our virtual machines that reside on a VMFS LUNS storage has asked if it is possible to have Non-VMFS LUNS presentable at our VM and VMFS?

  I think I remembered that Non-VMFS storage can be made available as storage but do not know if the two can be presented at the same time

  It can be done with a RDM and no special software required.

  -Matt

  VCP, vExpert, Unix Geek

• A Windows Server with direct access to a VMFS LUNS can climb the PowerCLI help?

  Hello

  I wonder... a Windows Server that has a direct access to VMFS SAN LUN cannot mount this partition using PowerCLI?

  Has anyone tried this before?

  Thank you

  -virtualhobbit

  No, that goes up the data store you specify (MyDS in the example, but you change this name).

  But the copy of a large VMDK on that provider is not a good idea.

  As you run the script on your PC, it will go on the network.

  As long as you have no functionality to copy at the end, an alternative is to copy the VMDK from the console with the ESXi vmkfstools command.

  See KB1027876.

  The problem is that you must be able to mount the destination locally on the node of ESXi to avoid going on the network.

• Please specify disadvantage of RDM, and how to modify vmfs to RDM lun?

  I read through many threads and still do not understand the disadvantage of RDM.

  #1 so, technically, the virtual machine is that a pointer to the SAN with RDM lun, correct? I could introduce this RDM drive easily to any server and see the data, and there may be a slight performance with RDM advantage.

  If this is the case, what is the advantage of VMFS to RDM with a SAN?

  #2 can a san LUN be divided into RDM disks will different VM?

  #3 if I have a unused 600GO VMFS san lun and want to make that a lun RDM, how should I do this? I need to fdisk and unformat the thing first? Then remove from storage and add?

  With the help of ESX 3.5 Update 2 and vCenter 2.5 Update 2.

  Thank you.

  #1 so, technically, the virtual machine is that a pointer to the SAN with RDM lun, correct? I could introduce this RDM drive easily to any server and see the data, and there may be a slight performance with RDM advantage.

  Fix.

  If this is the case, what is the advantage of VMFS to RDM with a SAN?

  VMFS advantages are that you can put a lot of virtual disks on a single volume.  Its a one-to-many relationship when the RDM is a one-to-one relationship to a logic unit number by computer virtual.  There is currently a maximum of 256 LUNS per ESX host, so if you had a large number of virtual machines you could get into trouble in this way.  3 RDM by 100 hosts would be a problem.  Another consideration is that the RDM add more it, the more you complicate the management of your storage environment.  Its much easier to manage a LUN 3 or 4 to 30-40 of them.  Usually the RDM are perfect when you need of MSCS, want to take advantage of the options of SAN backup, or if you have LUNS that must be transferred to a physical server at a given time.

  #2 can a san LUN be divided into RDM disks will different VM?

  I think that you do not want to break the LUNS as well.  Its much easier to manage these relationships 1-1.

  #3 if I have a unused 600GO VMFS san lun and want to make that a lun RDM, how should I do this? I need to fdisk and unformat the thing first? Then remove from storage and add?

  Not exactly, but I think you could remove the ESX host data store and then add it as an RDM to a virtual machine.  Fdisk should be able to fix it!  But back to the point 2 - it might make more logical management of skin only 600 GB and distribute according to needs.

  Good luck!

• A copy of the Lun in ESX 3.5

  Hi all

  Be friendly with me so here is my first message/question

  In my business, we would like to increase the size of the LUN as we are in a lack of space. We decided to do a copy-Lun to keep all the VM and reattached the lun/storage at the existing ESX (we did not want to manually copy all VM the former LUN to the new...)

  Admin doing storage copy (a test lun with 1-2 test VM on it...) then allowed us on our ESX, the function of "enablesignature". Rescan us the storage card (the new lun appeared...) then we have put back the storage on our ESX Server... BUT recreate the storage has asked us to "reformat" the vmfs partition (and we "lost" all VM inside...)  : 0.

  My question (and I know for sure someone already dealt with that...)...:

  How can we attach a copy of the old one (with VM already on it) to our ESX Server without losing all the data on it? I know that we could create an empty storage and move all the VMS from the old to the new... But because of my business processes, we have only one night to do and not enough time to move all THE VMS 'manually '...

  I forgot something? I did something wrong? After several attempts, same results: we've lost any datasbecause ESX that we were asked to formatt the new attached storage...

  Thanks in advance and waiting for your genius...

  Ford

  If the logic unit number are the same (because of a copy of low-level storage) your duty as a clone/snapshot (cause disk UUID is identical).

  You must use the DisallowSnapshotLUN = 0 option.

  http://www.yellow-bricks.com/2008/12/11/EnableResignature-andor-disallowsnapshotlun/

  André

  * If you found this device or any other answer useful please consider awarding points for correct or helpful answers

• Consigli sul compliant delle Lun

  Salve a tutti,

  nella EMC SAN ho a disposizione due nuovi cestelli di 15 brain in fibra da 400GO da utilizzare per macchine virtuali nel nuovo ambiente sleep 5 ESX.

  -Qual'e scelta migliore in questi casi finished delle performance e dell' use dello storage?

  -E' meglio creare aunt lun piccole o Pocket grew up?

  -E' meglio creare pochi grew datastore con più extended o più datastore con singolo measure?

  -RAID 5 o 10?

  Grazie

  Ciao,.

  TI consiglio di leggere questo articolo http://www.yellow-bricks.com/2011/07/29/vmfs-5-lun-sizing/

  Credo che possa fare al caso tuo.

  Personally cerco di mettere non più di 15 VM per lun, the lun sara given poi della dimensione dalla somma dei VMDK, più lo instant della più UN 20% di spazio libero per gli ram swap.

  UNA volta migrate VM nelle nuove monitoro che lo spazio libero non vada sotto lun what 20%

  CERCO inoltre di con delle lun, by far so che VM performance monitors ARE alte required di Ops / s no insistano lun sulla stessa.

  Ciao

• How can I format / erase a partition vmfs on the HARD drive for later use by windows?

  Hi all

  I have a HARD drive to a virtual host formatted as a vmfs partition. I need to delete / erase this HARD drive vmfs partition because it will now be used in a Windows environment. Can someone explain please how can I remove vmfs partition so that it can be formatted in NTFS? Can be done in a Windows operating system like the current place of residence may have one easily accessible Linux or vmware?

  Thank you.

  vmTrekker

  Windows should be able to reformat a VMFS partition in the GUI. If this isn't the case, you can run the diskpart utility and erase the partition table on the disk.

  André

• Block size is too small - reformat the data store or create files VMDk 2?

  Hello

  I'm new to VMware and so far I absolutely love everything about her!  Well, I was not too happy when I realized that I can't create more than 256 GB a drive without reformatting the VMFS datastore.   Apparently, I accepted the block size of 1 MB by default during the installation...   I now need to configure a file server with about 500 GB of storage.  The data store with a larger block size (4 MB would be fine for me) reformatting is easy enough?  I am currently working on the first host ESXi 4.1 and will move to set up 2 more hosts in the coming weeks.  I read somewhere on the forum that ESXi 4.0 doesn't let you change the block size by default of 1 MB - is this true or relevant to 4.1?

  Currently, I have only 2 small VMs on that host, and these can be easily saved and put offline for a few hours, if necessary.  When starting the server, I created a single RAID array that holds ESXi and all virtual machines - does that mean that I have to reinstall ESXi in order to increase the size of the block?

  An alternative that I can get to is to simply create 2 255 GB VMDK each and load distribution of storage in the guest operating system (Win 2008).   Performance wise, is a disc more great and better (or worse) than the two smaller disks?

  On a related note, what should I choose for the "Independent" option when you add a new virtual disk?  Default is disabled (not independent).

  Your thoughts, focus and expertise are welcome and will be greatly appreciated!

  Thanks in advance,

  Dothan

  Dothan,

  You should power (after the virtual machine backup) remove the current data store and create a new one with the new block size. I did not have this on ESXi 4.1 again, however, he worked on ESXi 4.0.

  The maximum size of VMDK. If you want ot be able to take pictures, make sure you subtract twice the size of block in GB of the documented maximum size.

  Block size 1 MB--> 254 GB (= 256 GB, 2 GB)

  Block size of 2 MB--> 508 GB (= 512 GB - 4 GB)

  Size of block 4 MB--> GB 1 016 (= 1 024 GB - 8 GB)

  Block size of 8 MB--> GB 2 032 (= 2 048 GB to 16 GB)

  André

  EDIT: Just the KB for the maximum sizes. http://KB.VMware.com/kb/1012384

• Formatting of disk/LUN

  Hello

  What is the best practice for formatting for VMFS disk/LUN. Currently, I have a 1. 6 t storage FSCSI. Formatting of disk/lun options

  The block size of 256 GB, 1 mb

  The block size of 512 GB, 2 mb

  The block size of 1024 GB, 4 mb

  The block size of 2048 GB, 8 MB

  From a performance perspective there is no real difference in the block size you are using.   From a disk waste point of view, there could be more wasted space when you use 8 MB block sizes.

  BP is to use the block size that fits the larger VMDK you will use on the VMFS volume.

  -MattG

  If you find this information useful, please give points to "correct" or "useful".

• Corrupt VMFS

  Hello

  for the error I installed esx 3.5 on a VMFS formatted LUNs, but with possibility to preserve the vmfs datastore.

  This is the situation:

  Disk/dev/sdb: 375.8 GB, 375809638400 bytes

  255 heads, 63 sectors/track, 45689 cylinders, total of 734003200 areas

  Units is 1 * 512 sectors is 512 bytes

  Device boot start end blocks Id system

  / dev/sdb1 * 63 208844 104391 83 Linux

  / dev/sdb2 208845 10442249 5116702 + 83 Linux

  FB sdb3 10442250 728612009 359084880 unknown

  / dev/sdb4 728612010 733993784 2690887 + f Win95 Ext'd (LBA)

  unknown/dev/sdb5 733801068 733993784 96358 + fc

  / dev/sdb6 728612136 729720494 554179 + 82 Linux swap

  / dev/sdb7 729720558 733801004 2040223 + 83 Linux

  the partition of fb is probably the VMFS data store...

  The data store is partially visible on some esx hosts... Are machine virtual working disconnected but VC...

  is there a way to recover the situation? to recover the vmfs partition? I see that all the other vmfs datastore depart from sector 128. can help remove the first 2 partitions?

  one way could be to remove the data store of rescanning vc and try and hba? should it be visible again?

  I can see the LUN on "storage adapters" on virtual center under its HBA.

  thx for the kind reviews...

  There is certainly a problem with "vmhba0:C0:T264:L0": seller: model 'IBM': "SAS SES-2 DEVICE" Rev: ' 02,0' ' no standard UID: failure '»

  See here: http://kb.vmware.com/selfservice/microsites/search.do?cmd=displayKC&externalId=9453805

  This will bring your store data, but under a different name and you will need to re - register your vm.

  -KjB

  VMware vExpert

• Expanding EMC Clariion LUN

  I would like to migrate LUNS using Navisphere manager to move a small logical unit number to a larger.  How can I get ESX to consider it as a ONE of the largest?  In Windows, you just see the new space using disk management and diskpart/expand to get the new space.

  Is this possible in ESX?

  No, you can't develop VMFS in this way. Only:

  (1) create new larger LUN, create new VMFS, migrate all virtual machines to new LUN, delete the old VMFS.

  (2) first save all virtual machines, remove VMFS, expand LUN, create VMFS on larger LUN, restore virtual machines

  (3) create new LUN on the place only where you want to add, add that LUN as a measure.

  ---

  VMware vExpert 2009

  http://blog.vadmin.ru

• VMware ESX 4.0 Update 1 with HP LeftHand P4500 Thin Provisioning (SAN based and client based?)

  I know with ESX 4.0 that you now have the option to select a provisioning on the side of ESX.  I'm already thin provisioning my LUN on the left side.  I do both or just the side SAN?  Also, I heard some people say that it is better to use a block size of 8MB for best peformance of SAN, do you think it is recommended/best practices as well?  Thanks in advance.

  The point to offer 8 MB block sizes is flexibility.  Because you cannot change the size of the block without reformatting the VMFS file system, you probably don't want to need to.  There is a great overview of a few points on this theme to the yellowwire from bricks.  I recommend you start here and follow the links.  It only takes 5-10 minutes, and you will be able to make your own choices.  For us, we're all migrate to blocks of 8 MB.

  What about thin provisioning, I have no practical experience.  But here's my point: If you sell thin anywhere, you have a more complicated system to follow in terms of space put into service and space required.  Must be taken account of this that data may coming in the future, or you might run out of space.  So, I ask you where you want to configure the monitoring of the use of space and how you want to administer the storage.  If you want to do on the primary server, and then stick with that.  If you want your VMware administrators to manage management, then do it in VMware.

  My recommendation: If you're already doing it on the back-end successfully, then stick with what you have.  Whatever it is, do it only in one place.  Otherwise, you're probably too complicate your configuration.

  Virtualization of happy!

  JP

  Please consider awarding points to useful or appropriate responses.

• Block the installation size

  Hello

  I have a situation where there is a new worm Server 4.0 build - local storage is 2.6 TB - I know that for any single player's local storage limits is 2 TB, so I divide it into 1 TB and 1.6 to - however, in our old 3.5 way of doing things, I would reformat the vmfs to a more appropriate to 1 MB block size.  Now, however, the service console is already put in the first vmfs.  So, the best practice is suitable for the size (a) by the first vmfs to accommodate only the supreme court, or (b) is at - it another way to change the size of the block I don't know?

  Thanks for any help

  Paul Flynn

  As much as I know there is currently no possibility to change the block size during installation or later. The best option would be to create a small disc on which the VMFS will reside for the Service console and create multiple or a single large (depending on size) LUN/Disk for other VMFS volumes.

  Duncan

  VMware communities user moderator | VCP | VCDX

  -

• Question ESXi 5.5 with turn a virtual computer: "an error was received from the ESX host turning on VM" '22 (Invalid argument)"'DiskEarly on a power Module failed.'

  Here's what happens when you try to turn it on one of my virtual machines (see also accessories):

  Error stack:

  An error was received from the ESX host turning on VM vzilla-ws2012r2e.

  Unable to start the virtual machine.

  Cannot open disk ' / vmfs/volumes/51286ca4-ef967828-664d-001b2129ad71/vzilla-ws2012r2e/vzilla-ws2012r2e_3.vmdk ' or one of the snapshot disks it depends on.

  22 (invalid argument)

  Power DiskEarly module has failed.

  Cannot open disk ' / vmfs/volumes/51286ca4-ef967828-664d-001b2129ad71/vzilla-ws2012r2e/vzilla-ws2012r2e_4.vmdk ' or one of the snapshot disks it depends on.

  22 (invalid argument)

  This circumstance may be linked to a sata cable issue, with the possibility of temporary loss of connectivity, which could result in data loss/corruption, I realize.  It is a laboratory box.  Especially say that the 2 VMDK he complains (trying to light) is both on the grounds of a single physical disk. Data, read and written to the speaker, since the problem are very good (which indicates the wiring problem was resolved, and the VMFS5 file system seems to be in good health).

  No photos.  No related clones.  Just a 2012 Windows Server based VM, with several drive letters in, with those underlying files VMDK residing on different stores of data VMFS5.  Implemented end (these drives aren't really so huge), but far from running out of physical space for the data either. Everything is working great for months, until today, trying to it turns on again.

  You are looking for:

  "Failure error disk beginning module lit" results in this article:

  error disk on start module failed

  which indicates the .lck files may be present.  It does not exist.

  Then upwards, of a variety of other items:

  Re: Unable to start the virtual machine: invalid argument on *-flat.vmdk

  http://KB.VMware.com/selfservice/microsites/search.do?language=en_US & cmd = displayKC & externalId = 1004232

  https://communities.VMware.com/message/search.jspa?peopleEnabled=true & userid = & ContainerType = & Container = & q = module + DiskEar...

  error disk on start module failed

  but alas, none of them seem to relate directly, or exactly.  My vmware.log file is attached below, as well as some screenshots of to show the structure of the unity of this virtual machine.  Hoping that this post proves fruitful, if anyone has had a similar situation.  The data at stake here are (mostly) redundant, but I would rather understand my way of it, in the case where it happens to me again, or can help others.  Many preferred rather than give up, reformat the VMFS and start again.

  Thank you!

  Good news, the best result I could hope. No data is lost. No corruption of the VMFS or NTFS don't drive in the virtual machine. Nice!  Saved me a few terabytes of data restoration and learned a little more on file system of troubleshooting along the way.

  It took a technician of VMware Service excellent, attentive, methodical remote 2 hours in a WebEx previously to resolve problems with these 2 files vmdk manually, because he found that there is a lock on them. I opened a request of Service (SR) # with VMware by following the instructions here:

  VMware KB: Cannot access certain files on a VMFS datastore

  To resolve this problem, apply for support from VMware Support and note this ID (1012036) Knowledge Base article in the description of the problem. For more information on the filing of a request for support, see How to submit a support request.

  I'll cover this saga and the exact process for collecting and downloading of newspapers, through to my TinkerTry.com, including the ride video.  I even captured much of the technical work that has been done. That said, it is true that little of the magic that was made to resurrect the metadata will remain a mystery, given that this piece is VMware.

  I'm ok with a bit of black box, considering how I'm happy that I got all my data, and time savings that the quick recovery represented.

  By clicking on the button "answer".

• VDP VMware data protection

  Hello

  I think to implement protection VDP 5.1 up to what my company approves more expensive and robust backup.

  There are instructions how to make and easy enough to follow, but one thing remains unclear. There are 3 types of 0.5 CT 1 to and 2 TB, which indicates the size of the disk for different flavors. to 0.5 to 1 TB things are clear because I can add in 5 VMFS formatted LUNS and get 2 TB of space but what happens with 2 TB POS. 2 TB VDP will be 3.1 TB of space.

  My question is what type of storage I must add to ESXi host to have for example 4 to LUN. What I need to add NFS storage and install the unit on the NFS storage.

  My hypothesis is that 5 VMFS datastore maximum size is 2 TB which means that 2 TB POS does not fit into here.

  Is there a way to add 4 TB of storage without formatting as store VMFS which be limited to 2 TB. For example, I add 4 TB of storage introduce him to ESXi host and install the device

  Maximum size of VMFS datastore is not 2 TB. His 64 to. You can create up to 64 TB LUn and add it as a single data store. So, you can easily create 4 TB LUN and make a data store for your POS.

  Check this box

  VMware KB: Frequently asked Questions on VMware vSphere 5.x for VMFS-5