hi,
i have two 3Tb disks in raid 0 on regular SATA ports but esxi shows them as two separate disks
a similar raid on my sas 2308 controller does show the correct disk information through a "ls /vmfs/devices/disks"
my sata controller is defined by Vsphere as a "Wellsburg raid controller"
the (boot) ssd disk on another sata port of my esxi server does work fine.
anything i can do to fix this or research further?
Hi,
I am having a strange issue that I can't figure out. We are (constantly) in the process of upgrading our old hardware which includes VMware hosts. Every time I've upgraded I usually get the new hardware and cold migrate the VMs over then decommission the old hardware. After at least 30 replacements I came to the one that is having the issue. I have the old host which is an IBM x3650 M3 running ESXi 5.1 (2000251) and the new/replacement host which is a Lenovo X3650 M5 running ESXi 5.5 (2718055) I have had the same exact builds running on an old and replacement server in other locations.
Now the problem I'm having is both hosts are plugged into the same gigabit switch with CAT6 cables. There is no shared storage, they are both managed by vCenter with Standard licenses. The switch is configured correctly and anywhere I can think to look all of the connectivity shows it is at 1gig. Both hosts use 2 NICs on 1 vSwitch with load balancing. They each have one virtual machine port group and 2 VMKernel ports, one for management and one for vMotion.This is the same setup I have at every one of our branch office locations.
If I transfer files or use iperf between a VM on each of the hosts I get gigabit speeds. When I try to actually migrate my VMs from the old host to the new I get 100mbps with cold migration or vMotion. I have never run into this before and I am lost on where to look next. I combed through the switch with our network admin and that looks good. I'm assuming since I do get gigabit between hosts via the VMs that the network is ruled out anyway.
If anyone can point me in any direction at all I would appreciate it. I talked with VMware and they weren't any help. I will probably end up calling again but I'm hoping someone here might have some ideas. I tested iperf between the VMs on port 443 as I believe that is what the cold migration uses (management port).
Any help is greatly appreciated.
Hi all,
I had this PSOD after enabling the PassThru to a HP SAS HP H241 on HP DL380P G9 with ESXi 5.5 U3b.
2016-02-25T15:16:41.507Z cpu8:32851)@BlueScreen: #PF Exception 14 in world 32851:helper1-0 IP 0x41800b0bc1c8 addr 0xf8
PTEs:0x123e42027;0x8955b5027;0x0;
2016-02-25T15:16:41.507Z cpu8:32851)Code start: 0x41800b000000 VMK uptime: 0:00:34:27.461
2016-02-25T15:16:41.508Z cpu8:32851)0x4123814dde50:[0x41800b0bc1c8]VMKPCIPassthru_GetOwnerVmmWorldID@vmkernel#nover+0x8 stack: 0xffffff
2016-02-25T15:16:41.508Z cpu8:32851)0x4123814ddf00:[0x41800b06994c]IOMMUDoReportFault@vmkernel#nover+0x16c stack: 0x200000800
2016-02-25T15:16:41.508Z cpu8:32851)0x4123814ddf30:[0x41800b069ba7]IOMMUProcessFaults@vmkernel#nover+0x1f stack: 0x0
2016-02-25T15:16:41.509Z cpu8:32851)0x4123814ddfd0:[0x41800b061455]helpFunc@vmkernel#nover+0x6a1 stack: 0x0
2016-02-25T15:16:41.509Z cpu8:32851)0x4123814ddff0:[0x41800b256f9a]CpuSched_StartWorld@vmkernel#nover+0xfa stack: 0x0
2016-02-25T15:16:41.515Z cpu8:32851)base fs=0x0 gs=0x418042000000 Kgs=0x0
I had a workaround ==> https://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=2032981
But I need a fix to avoid the PSOD.
Any help ?
when installing patches manually, is there a way just to install specific VIBs
my co-worker has installed the following VIBs on top of a vendor provided esxi 5.5 U3 build. I want to install the same VIBs on another host. The esx base looks easy, as it is the only VIB in the patch. But what about the lsi VIBs, they are part of a patch containing many VIBS.
lsi-mr3 0.255.03.01-2vmw.550.3.68.3029944
lsi-msgpt3 00.255.03.03-2vmw.550.3.78.3248547
esx-base 5.5.0-3.84.3568722
I am familiar with the command # esxcli software vib update -d "/vmfs/volumes/Datastore/DirectoryName/PatchName.zip" .
-Some of the VIBs will be skipped as they are already present version or newer. But I am concerned about VIBs in the patch that are not present, and dont want the patch to add them. Can I specifically install just the lsi VIBs from a command?
So, I have two ESXi5.5 hosts connected to some iSCSI storage. Both hosts are identical in every respect, inc the dual-port Intel X540-T2 10GbE NIC in the same slots. The storage array has two dual-port X540-T2 NICs. I'm connecting the hosts to the storage using CAT6A x-over cables. The hosts are using the standard ESXi software iSCSI adapter. There are no errors on either host. and the vCenter Appliance is happily talking to both hosts with HA/DRS enabled.
I needed to performance tune the iSCSI connection to the storage, and along with some basic modifications to the iSCSI S/W adapter on each host to match the storage array, I wanted to switch to using 9000-byte jumbo frames.
I made the jumbo frame setting changes to the storage, then made the changes to the ESXi hosts. The first host went perfect - a quick rescan of the HBA's, and the storage was working fine - much quicker as well. Then I made the IDENTICAL change to the second host, and it went into a complete fit. First, it completely lost connection to the storage, then I couldn't view any Network Adapter, Storage or Storage Adapter settings. The config pages wouldn't load, and I eventually lost connection from the VIClient. SSH still seemed to work, but no commands would be accepted. The console was also unresponsive, so I had to do a hard reboot. The host started to boot, but was VERY slow when scanning for iSCSI volumes. It did finish booting after about 10 minutes though.
At this stage, I could get back into the config settings, but the iSCSI volumes weren't showing, but the iSCSI LUN's were showing as available in the host iSCSI HBA! I tried to force mount the volumes, and the host went into meltdown again (same symptoms). Another hard boot. After about an hour of fighting with this, I changed the MTU back to 1500 on the vSwitch connected to the storage array. Almost instantly, the storage re-appeared and re-mounted. Change MTU back to 9000, same fit. I triple checked everything - all settings looked ok. I switched the host storage connection to an unused 10GbE port on the storage array - same problem. I changed the x-over cable - same problem. Nothing I could do would make the host use a 9000-byte MTU, but the other host is happily using it!
Beats me. Got to try and get this sorted before this system goes into production.
Anyone got any ideas?
Hi all,
Due to financial constraints I am going to use a nested solution for my home lab on a powerful white box machine with at least 32 GB of RAM and Xeon processor but I need some clarification and I hope someone will answer my questions shown below:
1. Does it matter what motherboard I get if my home lab to be nested on VMware workstation?
2. Is the VMware Compatibility Guide for bare-metal hypervisor?
3. Which of these motherboards (ASRock, ASUS, GIGABYTE, Intel, MSI and SUPERMICRO) is suitable for my home lab?
Thank you and I look forward to hear your recommendations?
Cheers, Nooga
Hi all,
I have ESXi 5.5 on VMware worstation which installed on linux.
I coverted .vmdk to .qcow2 disk images and create new VM on KVM.
VM successfull run, but network insterface doesn't work. It was finded in system, but always write me - that it down
esxcli network nic list
Name vmnic0
Driver e1000
Link Down
Could you help me please to solve this problem?
what is wrong ?
how to change link to up?
One of our customers uses an ESX 5.5 VMware system to host a Windows Server 2012 R2 Standard in which we run our software. The VMware Tools version is 9.4.5 (build-1598834).
We notice that after running for weeks the amount of used memory in the hosted system steadily increases: the non-paged pool memory increases to many gigabytes (non-paged pool memory under normal circumstances should only be a few hundred megabytes).
After an analysis using the Windows poolmon tool it shows that the majority of this memory is used by a "VNet" driver which seems to come from VMware.
Is this a known issue? What do you recommend?
Good day!
I'm using free version of your VMware vSphere Hypervisor 5.5 installed on a flash-drive. Do you have any solution or maybe best practice manual to make backup all the information from ESXi-host (include config, OS, etc)?
Thank you in advance.
I am due to do some vmware patching and network driver installations. I read on a blog that it is a good idea to backup up the configuration using , as shown at VMware KB: Backing up and restoring ESXi configuration using the vSphere Command-Line Interface and vSphere PowerCLI . , using the vim-cmd hostsvc/firmware/backup_config command, to create a configBundle-HostFQDN.tgz
-I am not clear on what scenarios this type of backup is useful for. Is it suitable for recovering from patch or driver install issues?
-I note that bootbank information is not stored in this type of backup, would I need to backup that too in a driver /patching scenario?
-for restoring, I guess its just a case of copying the backup file to a datastore and running the command e.g vim-cmd hostsvc/firmware/restore_config /tmp/configBundle.tgz?
Dell have asked me to update FCOE drivers, due to an intermittent lost connection to datastore message.
How are the drivers updated? The vmware page at VMware Compatibility Guide - I/O Device Search suggests bnx2fc version 1.712.30.v55.27.12 is compatiblewith my firmware Qlogic BCM57xx is 7.12.19.
Do I just download the driver at VMware vSphere 5: Private Cloud Computing, Server and Data Center Virtualization , and install the package using the: esxcli software vib update -d "/vmfs/volumes/datastore1/patch-directory/ QLG-NetXtremeII-3.0-2634798.zip command? I note the driver will update multiple vibs, even those not related to FCOE, I guess it makes sense to update them as a set? Then reboot after?
Hi guys.
I'm having an issue, when I create a Disk with 500GB or more and try restore a backup of my SQL I got a speed of 6MB/s.
When I create a disk with 300GB less, my speed to restore backup is 100mb/s
Someone can tell me why occur this problem?
Thanks
Diego Rodrigues
Hi guys,
I've read a lot of discussion all around the Internet, mostly found some issues regarding LSI MegaRAID SAS 9240-8i, which differs from mine card just by having 2 mini-SAS ports instead of one, driver for this controller should be the same.
In ESXi 5.0 there was some workaround handling with ESXi Customization, so the driver should be placed into the original ISO file. In ESXi 5.0 U1 should be driver for LSI MegaRAID SAS 9240-4i/LSI MegaRAID SAS 9240-8i natively, but it still doesn't work for me.
Important information is when controller is not plugged into PCI-e x16 port, everything works fine (ESXi Installer will pass and ESXi is running from selected device).
My problem is: When trying to boot ESXi Installer (5.0/5.0 U1) everytime I got stucked when drivers are beeing loaded, namely "megaraid_sas". Screen looks like this:
Anyone can give me some serious help?
Thanks in advance,
Jakub Rybar (Kubicz)
I need to list all host's install date, is there a way to determine host install date.
I have recently installed VMware ESXi 5.0.0 (release build 469512) on Intel DQ67SW motherboard. After the reboot, ESXi is not booting up automatically even the first boot device is HDD. I am getting following error message
A BOOTABLE DEVICE HAS NOT BEEN DETECTED.
PLEASE REFER TO THE PRODUCT GUIDE AT HTTP://SUPPORT.INTEL.COM/SUPPORT/MOTHERBOARDS/DESKTOP
Then I have to give CTRL+ALT+DEL and F10 for booting from HDD.
.
Processor: INTEL I5 2500
MB: DQ67SW
SATA HDD: ST3320418AS
Please let me know if any have idea ?
Thanks
Mahamood
I need to find the right way to confngure VMs hosted by small ESXi hosts.
Given a small ESXi host (let's say, 1-2 physical CPU, 4-32 cores, 16-48 GB RAM) and given 2 guest Windows VMs, what is the right way to assign resources to host VMsif they are supposed to run concurrently?
If I have only 1 physical CPU is it advisable to assign no more than 1 virtual socket to each host VM?
If I have 2 physical CPUs, is it advisable to assign 1 or 2 virtual socket to each gueast VM?
Should I keep the overall number of cores assigned to guest VMs under the number of physical cores?
How much memory can I assign to guest VMs?
Should I reserve some RAM for the ESXi host itself or can I assign all the available RAM to the guest VMs (let's say, if I have 16 GB RAM I can assign 8 GB RAM to each one of my 2 guest VMs)?
Is there any document I can use to answer my questions?
Regards
marius
Hi,
I need to verify whether configured scratch partition is working fine or not and also has to verify logs are getting updated.
Can some one please suggest some ways to do this. If PowerCLI script also fine.
Thanks in advance....
Imagine a vSphere 5.5 infrastructure based on some ESXi v 5.5 hosts.
There is a standard virtual switch spanning all the hosts.
I need to collect data (in terms of percentage of bandwidth used) about the traffic from given VMs to a specific host connected to the external network.
Let me give a simple example based on a simplified scenario:
vSwitch1 is a standard virtual switch and is connected to the physical 1Gb NICs of ESX1, ESX2 and ESX3.
All the 3 1Gb NICs are connected to a physical switch
A physical server (let's say, Server1) is connected to the same physical switch as the 3 ESXi hosts.
I need to monitor the amount of data transferred between VM1 and Server1, VM2 and Server1, VM3 and Server1 to see how they use the theoretical 1Gb bandwidth they could use concurrently (given that the bottleneck is the physical connection between Server1 and the physical switch)..
I see I can monitor in realtime the amount of traffic for each physical NIC of each ESXi host: can I do anything at the vSwitch side?
Can I collect data during a given timeframe?
Regards
marius
Dear All,
I was doing testing for deploying a virtual machine from a template, the template is created from a virtual machine which was having few files on desktop in 'administrator' profile. the same files which are on desktop are also in C:\ drive of original machine.
After I deployed a new virtual machine from template I found same files as original machine on desktop and C:\ drive but now I wanted to change SID of new machine so I ran sysprep from Windows> system32 \ sysprep, but after the process completed I found on the new machine that files from desktop are gone while same files were still existing in C:\ and also applications were there so what does this mean ? sysprep clears desktop items ?
Today I logged into vSphere and this notification came up.
Lost connectivity to the device mpx.vmhba32:C0:T0:L0 backing the boot filesystem /vmfs/devices/disk/mpx.vmhba32:C0:T0:L0. As a result, host configuration changes will not be saved to persistent storage.
And in the event log as well:
Bootbank cannot be found at path '/bootbank'
I did some research and it suggested that my USB drive (on which the ESXi was installed) had failed. Yes I know there are a gazillion of posts on this issue but I believe mine is a bit different:
So, I did not quite believe my USB drive died so fast (not even 1.5 years and I am not a heavy user), so I pulled it out and put it into my computer, and 3 partitions came up, the USB drive was successfully recognized.
Then I put the USB drive back, SSH into the ESXi and "lsusb", my USB drive showed up as "Bus 02 Device 0b: ID 0781:5580 SanDisk Corp. SDCZ80 Flash Drive".
Now, if I am to play it safely, my best bet would be to replace that USB drive. However, I am not sure if my USB drive has actually failed?
Update: It does show "Input/output error" when I try to navigate into /bootbank in SSH.
/vmfs/volumes # ls
ls: ./79a964e1-88c0598a-7b89-b1a18986d360: Input/output error
ls: ./28b21ce3-3914b9a5-1fea-afb54459361e: Input/output error
ls: ./54b1487c-dce8d390-11e2-0cc47a455f8e: Input/output error
I am still a bit skeptical about whether my USB drive has actually failed. I am going to install ESXi on another USB drive and try it.