ReadyDATA and VAAI

Similar Questions

• Controllers of the array failover (Dell md3020i, esxi5) and VAAI

  We lack 6 5.1 esxi hosts that use the vmfs5 storage provided via iscsi from a double controller Dell md3020i.  Our charge approximately 900 inputs/outputs, but can boost 1800 IOPS / s.

  During periods of high load, our virtual disks to failover from one controller to another.  This does not stop time, but produced a few problems on the virtual disk of balancing on the controllers.  We have opened a folder with Dell, and they see mistakes ancient IO from esxi hosts.  Dell is investigating a possible firmware (not yet released) was updated and also suggesting an update of firmware of the hard drive.  They also mentioned that we can almost be the ability of the matrix (in terms of IOPS / s).

  We also opened a case of VMware support.  They recommend that we disable VAAI. (http://kb.vmware.com/selfservice/microsites/search.do?language=en_US & cmd = displayKC & externalId = 1033665).  But all what I have read talks about the benefits of performance of the VAAI and our table officially support it.  I ran this suggestion by Dell, and they disagree - recommend us keep active VAAI.

  So, we are stuck with what to do next.  Suggestions from the community on VAAI or other changes to help with our table failover problem?

  Thank you

  I don't think that VAAI is a problem. It could due to an overload on the disk due to the IOs, it is moving to the other controller. Please update the firmware...

  You have excerpts from newspapers pointing to older IOs that shows problems with ESXi

• ESXi 5.0/5.1 and VAAI Compellent

  I can't try to understand VMware 5.0/5.1 with a Compellent SAN Setup. I am trying to collect information on the current configuration of the VAAI but couldn't get anything to go up.

  Command:

  List of devices esxcli vaai

  Result:

  Error: Unknown command or namespace vaai list of devices

  It seems that the game complete command vaai is either not installed or is not available and I have no idea why. I have several hosts ESXi 5.0.0 U1, 5.1, updated since the previous running and cleaning 5.1 installs. I'm under 5.1, 5.1, 5.1 PowerCLI vCLI vCenter. I tried direct console as well as the remote shell connections.

  Is there a package that I need to install on the host computer to activate the command VAAI set? Is there a service that I need to start?

  To the Compellent SAN is running firmware version 5.5.3 soon to be upgraded to 6.x - Compellent Enterprise Manager is currently 6.2.2.8 running and my VMs using physical RDMs have installed the Agent last Compellent. My main goal is to allow the recovery of space Compellent that I have read and learned should be supported in my configuration.

  Any help would be appreciated to determine issues of VAAI namespace to start.

  Thank you

  Scott K.

  You are using version 4.1 of the order.

  Try this instead:

  esxcli peripheral central storage vaai situation get

• VAAI & ATS data warehouses

  Hello

  just a small Question. If I created and formatted a Datastore before VMFS5 SAN Bay was able VAAI data store after a Firmware Upgrade, now supported VAAI, will use ATS?

  If I create a data store on a VAAI select Bay SAN is the ATS Datasore only... so something to do here?

  Greetz,

  Peter

  Yes if you updated the firmware and VAAI is supported should be automatically enabled after a storage operation 'Refresh', you can check by following the instructions in the following link: http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1021976 : ATS only it will not use SCSI-2 reservations, this is not a mistake and should on VMFS-5 volumes again. (see the link for more information).  You can disable individual VAAI fundamental operations, use of advanced settings if you need to.

• SCSI reservations esx 5

  Hello

  We have a project with 48x900gb sas hus.migration hitachi virtual machines of the old storage from hitachi.

  There are 100 computers a virtual especially windows 7, some esx host of linux.5 by fibre channel switch and storage and ISL connected with other 5 hosts and other storage containers.

  Because the disc is too big I think for a mix of raid 5 7 + 1 and 2 raid 10 raid in 2 pools of hitachi groups, large stripping.

  but when I present too big lun in esx will be a problem of reserves scsi? or is it only when I have too much intensive vms in a data store? I want to keep 1 lun = 2 data warehouses.

  Hitachi has 2 LUNS of controllers.all I have will present 2 ports to each controller or is better one port per controller? each esx should see tha luns.is even it correct as implementation? its my first project and I don't have too much experience.

  johnxgr wrote:

  I wanted to say the following.

  due to the large drives and when I'll build raid 5 lun will be big.generally will be a problem to present to esx 5 TB as a data store?

  ESXi 5 it can withstand without problem.

  or is better to split in two?

  Possibility - you should consider that a single LUN has only a single queue, which may present a problem of system.

  make sense that due to scsi reservations?

  There is no reserve of SCSI.

  In rule General when I shared a lun and in each one I have VM the queue is one.for perfomance reasons plays no role to divide?

  Queues absolutely make a difference.

  and vaai in hitachi works when I have a pool of raid groups and with the selection broad stripping activated?

  You must ask Hitachi who.

• Size of data store on smaller clusters

  I have a bunch of data stores in the range from 500 to 750 GB on a 3 host cluster. The SAN is a group of equallogic, which designates a raid by box group. With the latest firmware installed it is support VAAI.

  Is there a reason for me at all do not use warehouses of data of size max? I don't see an issue with about 50 VM and VAAI scsi reservations.

  -Anders

  You will have a depth of queue device and multiple LUNS means also several queues indeed. Even if it's the same raidgroup you will avoid always fills a single queue and hurt all your virtual machines.

  Duncan (VCDX)

  Available now on Amazon: vSphere 4.1 HA and DRS technical deepdive

• Reference Dell R710, Dell and VMware VAAI MD3200

  Hi guys, we have 3 x Dell R710 hosts running VMware ESXi hypervisor v5.1. Our SAN is via Dell MD3200 Direct attached storage. We have activated the option of High Performance. My question is about VMware VAAI. In particular, the recommendation of VMware: http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1033665 it says: "For more information on support VAAI in a particular storage group or required firmware levels, contact the storage array provider." So I'm wondering: is VAAI supported (or not) with the MD3200... and otherwise: should I be deactivation of VAAI settings as recommended by VMware? What are the potential implications for the performance of deactivation of VAAI... or support DO NOT turn it off? Thank you, Steve

  Indeed, you are using a fairly old firmware. The most recent is 07.84.53.60.

  To be able to upgrade to this version, you will need a version quite familiar of the MDSM. I would say to download 4.1.2.29 or later, which is 5.0.0.70. All servers Windows and Linux (not VMware or XenServer) server that manage and/or are connected via iSCSI SAN should have their MDSM, updated software before upgrade you the firmware on the raid on the SAN controller.

  RAID controller firmware: www.dell.com/.../powervault-md3200

  Latest resources DVD (5.0.0.70): www.dell.com/.../powervault-md3200

• VMware VAAI and SCWIST

  I tried to configure VMware with iSCSI running on Ubuntu Linux SCWIST target, but I couldn't order EXTENDED COPY working between LUNS. It works correctly in SCWIST with sg_xcopy command but in VMware it doesn't. The EXTENDED COPY command only works in the same LUN. I analyzed the traffic with Wireshark and noticed that the COPY of the SCOPE is executed when the virtual machine is cloned in a data store. I discovered that NexentaStor announces that it works with VMware VAAI, so I tested it with their free trial. The NexentaStor was unstable, but sent VMware EXTENDED COPY commands between the LUNS (data warehouses). I captured the session of connection set of NexentaStor with Wireshark and modified code SCWIST to respond with the same data as NexentaStor (with the exception of the seller, identification of the product in standard INVESTIGATION command). It did not work. The last step was the development of the fields of identification (IDENTIFICATION of the SELLER and PRODUCT IDENTIFICATION, PRODUCT REVISION LEVEL) in the standard SURVEY and... He started working. I back return all changes, leaving only these three fields which introduce the reader that one of NexentaStor and VMware did not discover the iSCSI LUN. So I also made a similar identification device VPD page to of NexentaStor. Does that mean containing NAA indicator, indicator specific seller target port, indicator for the target port group and indicator of relative target port. Devices have been discovered then and order EXTENDED COPY was sent during cloning machines between data warehouses (LUN).

  How is it possible that VMware has hardcoded behavior of the target of NexentaStor? This isn't even a VAAI plugin:

  esxcli peripheral central storage vaai situation get naa.606144f0e0d4fd770000503970705800 d

  VAAI Plugin name:

  ATS status: supported

  Clone status: supported

  Zero status: supported

  Delete status: supported

  I would use VMware solutions with open source and I expect that it includes in the same way that when there is no plugin owner.

  The SURVEY sample response that makes the works similar to below:

  INVESTIGATION of standard:

  PQual = Device_type RMB = 0 LU_CONG = 0 version 0 = 0 = 0 x 06 [SPC-4]

  [CREA = 0] [TrmTsk = 0] NormACA = 0 HiSUP = 0 Resp_data_format = 2

  CSC = 0 VAC TPGS 3PC = 1 0 = 1 = protect = 0 [BQue = 0]

  EncServ = number 0 = 1 (VS = 0) [MChngr = 0] [ACKREQQ = 0] Addr16 = 0

  [RelAdr = 0] WBus16 = Sync 0 = 0 [related = 0] [EPP = 0] CmdQue = 1

  [SPI: Clocking QAS 0 x 0 = = 0 0 = IUS]

  length = 66 (0x42) device type: disk

  Identification of the seller: NEXENTA

  Product identification: COMSTAR

  Product revision level: 1.0

  Unit number: 3afcb739

  Peripheral Identification VPD page:

  Addressed logical unit:

  indicator type: NAA, code set: binary

  0x606144f0e0d4fd770000503970705800

  Target port:

  indicator type: specific provider [0x0], code set: ASCII

  transport: Internet SCSI (iSCSI)

  vendor specific: iqn.2016 - 05:com.testtg0

  indicator type: Group of target code set port: binary

  Target port group: 0x0

  indicator type: Relative, code set target port: binary

  Relative target port: 0x1

  Isaac Goldbaum

  It only works with the constant IDENTIFICATION of THE PRODUCT field. Whatever it is.

  Isaac Goldbaum

• VAAI - what I mean... and what don't get me?

  So I am running vSphere 5.5, and my data stores reside on a HP 3PAR 7200, using the 10Gb iSCSI Protocol. In order to recover space on my thinly provisioned volumes, I run "esxcli storage vmfs unmap-l < name >". It seems free storage for allows me to shrink my GPC and space of my 3PAR recovery. watching vSphere STANDARD 5.5, which I... looks like my license does not include VAAI... so I have runs a command on my ESXi hosts that actually does nothing? What exactly I don't go as far as integration of storage with only the standard license and if I don't get to cancel mapping, why he accepts the order... and seem to free space on my Thin configured virtual devices? I'm confused. Also, if I don't get the ability to send cancel the MAPPING as part of my license, how am I supposed to pick up the storage?

  I think that instead of "unloading" it you're host resources will used beeing could have the VAAI operate.

  the only Point that is explain for me.

  So at the end would be that performance is better with Enterprise > due to unload all orders

  But I couldn't guarantee my "theory"... It would be good to open a Ticket to vmware and ask this question.

  Best regards

  Max

• Confusion on table VAAI and EMC

  I have a question regarding the use of the VAAI and an EMC VNX. So, since the host command line you can check the status of vaai and it shows different vaai components that are supported. Just because they are caught in charge doesn't want to say vaai works, right? I saw some documentation showing the use of AUVS to install the plugin vaai, I wasn't sure how all this works. I guess with all the tables, if they support vaai to install the plugin. It is not something that is just Bay storage aware, is it?

  Generally, you don't have any special plugins/VIBs side ESXi operate VAAI with specific storage devices. This is all implemented SCSI standard controls that the storage array is supported or not.

  I have a question regarding the use of the VAAI and an EMC VNX. So, since the host command line you can check the status of vaai and it shows different vaai components that are supported. Just because they are caught in charge doesn't want to say vaai works, right?

  If the output of get esxcli central storage device vaai situation indicates that VAAI is supported for ATS/Clone/Zero operations (Delete a bit is another story), so you can safely assume THAT VAAI operates.

  To confirm this, you can check esxtop (r) while you clone a virtual machine or create a disc of eagerzeroedthick:

  vmkfstools 5-G - c-d eagerzeroedthick /vmfs/volumes/MY_VAAI_DATASTORE/testdisk.vmdk

  While checking the view of storage (r) esxtop ('u') device and add fields for the display of stats VAAI ('f', 'o'), remove other fields if it does not fit on your screen. While the disc is being cancelled, you should see that MBZERO/s for the LUN is high and the normal MBWRTN/s meter shows no significant activity.

  If VAAI is not actively used, the command vmkfstools above creation of the disc will result in normal writing not discharged and the counter values will flip over.

  Also know that for copying/cloning between VAAI capable LUNS, source and destination must be the same version major VMFS and blocksize, and two LUNS must be managed by the same storage controllers. Otherwise, it will automatically change back to not unloaded read/write operations on LUNS.

• Edition VAAI and Essentials kit

  Hello

  Looking at the different editions of vSphere obviously support VAAI is compatible that with Enterprise and Enterprise plus editions. This means that when you use another edition (Essentials in my case), I have to disable VAAI in Esxi hosts, although the storage array can support? And besides it's meaningless to activate VAAI in the host and the table since I was on the edition of Essentials?

  Thank you

  If your license does not allow for it, it will not be used, regardless of what that announcement Board.

• VAAI and VM VDisks performance

  Hello

  Outside operations VM as cloning, migration management, creating vDisks, etc... should VAAI also affect performance once the virtual computer is created and configured?

  In my case, I am conducting a test using Iometer in a virtual machine with different configurations of virtual end/thick disks, ignition and extinction VAAI and I realized that there is no impact on the results. I wonder if this is the expected result.

  Its actually probably because the difference in performance of a thin disk versus a Recursive drive is very very small (below 2%) and therefore very difficult to measure.

• IBM midsize v7000, VAAI and storage plugin

  I'm trying to understand what kind of support does the IBM midsize v7000 to NOTTM.

  It seems that only the primitive VAAI 'zero block' is supported, but it requires the vmw_vaaip_t10 plugin to do this.

  Apparently, even though I am familiar with storage and vmware I never need to use a plugin for different than default storage, even if I saw it several times when I change the path failover strategy I have never studied on how to change it.

  Can someone confirm my conclusions on IBM midsize v7000?

  If they are correct, do you know if supporting the other two primitives of VAAI in the roadmap?

  Can someone tell how can I change the plugin to mw_vaaip_t10 and if and how this has effects on the strategy of path failover?

  Thank you in advance.

  Primitive VAAI 'Reduction to zero of the blocks' is supported with vmw_vaaip_t10 plugin.

  I have no idea of the roadmap VAAI

  To configure NOTTM for IBM statements using the VMW_VAAI_T10 plug-in, use the following sequence of commands:
  # esxcli corestorage claimrule add - claimrule-class = filter - plugin = VAAI_FILTER - type = vendor - seller = IBM - autoassign
  # esxcli corestorage claimrule add - claimrule-class = VAAI - plugin = VMW_VAAI_T10 - type = vendor - seller = IBM - autoassign
  # esxcli corestorage claimrule - claimrule-load class = filter
  # esxcli corestorage claimrule - claimrule-class of load = VAAI
  # esxcli corestorage claimrule run - claimrule-class = filter

  This is not an effect on the path failover policy.

  Do it me know that responds to your request or you need more information?

  Kind regards

  Satya

• ESX 4.1 VAAI and IBM XIV 10.2.4... does not work?

  Hello

  We had our set XIV to level to the last code 10.2.4 allowing VAAI support. We followed everything that the http://kb.vmware.com/selfservice/microsites/search.do?language=en_US & cmd = displayKC & externalId = 1021976 ... but still our hardware acceleration info shows is unknown or not supported. We manage the ESX 4.1 u1 so we have the VIB:s last support VAAI etc.

  If take a look in an ESX on the ATAS plugins available running

  esxcli corestorage claimrule list - ch. VAAI

  We get this result:

  Rule class class Type Plugin Matches

  VAAI DURATION 65430 VMW_VAAIP_SYMM provider = model = SYMMETRIX EMC

  VAAI file 65430 VMW_VAAIP_SYMM provider = model = SYMMETRIX EMC

  VAAI 65431 TERM VMW_VAAIP_CX provider model = DGC = *.

  VAAI 65431 file VMW_VAAIP_CX provider model = DGC = *.

  VAAI 65432 DURATION provider model = EQLOGIC VMW_VAAIP_EQL = *.

  VAAI 65432 file provider another VMW_VAAIP_EQL model = EQLOGIC = *.

  VAAI 65433 DURATION provider NetApp = VMW_VAAIP_NETAPP = *.

  VAAI 65433 file provider NetApp = VMW_VAAIP_NETAPP = *.

  VAAI 65434 DURATION supplier HITACHI model = VMW_VAAIP_HDS = *.

  VAAI 65434 file provider model HITACHI = VMW_VAAIP_HDS = *.

  VAAI 65435 DURATION provider model = LEFT VMW_VAAIP_LHN = *.

  VAAI 65435 file provider model = LEFT VMW_VAAIP_LHN = *.

  and information on a data store, it looks like this: (esxcfg-scsidevs - l)

  EUI.0017380005e200d2
  Device type: Direct access
  Size: 966656 MB
  Full name: IBM Fibre Channel disk (eui.0017380005e200d2)
  Multichannel plugin: NMP
  The console device: / dev/sds
  Devfs Path: /vmfs/devices/disks/eui.0017380005e200d2
  Vendor: IBM Model: 2810XIV Revis: 10.2
  SCSI level: 5 nickname is: wrong status: on
  RDM Capable is: true is removable: false
  Is Local: false
  Other names:
  VML.01000800003630303035453230304432323831305849
  VAAI status: unknown

  Let him know that it is a XIV and his saw: 10.2 on?

  Do we not have to manually install the plugin ATAS vendor: IBM model: XIV? I read a quick guide for 3Par, which showed how to their storage...

  Does anyone have a good answer for this?

  And as I said before... We followed the F.A.Q for this so that the size of the block etc, it is the same for all data warehouses and so on.

  grateful for answers.

  BR

  Henrik

  You are right. I got the message of support XIV support VAAI was held for a few weeks because of a problem with it.

  It would also mean that there will be a 10.2.4a or even firmware. We had improvements planned for February 25, but was asked to wait until March 15 for an updated firmware.

• ReadyDATA of ReadyNAS data import

  We are trying to deploy a ReadyDATA 5200 and refocus our ReadyNAS 4200. Is there a quick way to transfer data or actions of a ReadyNAS to the ReadyDATA. I looked at the link of imports by volume, but this seems to be only for import data from the seed.

  Yes, the import function is just to import seeds for ReadyDATA jobs repeated data.

  I recommend you create the necessary actions on the ReadyDATA. Enable SMB or NFS on the shares.

  Then create on the ReadyNAS backup jobs using the ReadyNAS built-in Backup Manager. Create a job for each action and push data across the ReadyDATA by using SMB or NFS.