[Home] [Forum]

The page contains a number of common issues that can be encountered running ESXi. If you have any questions about these tips, please post a question in the forum. If you have an issue and solution to add, use the comment form below.

1) During an install of ESXi you encounter the errror "Unable to find a supported device to write the VMware ESX Server 3i 3.5.0 image to"

2) After an install of ESXi, you reboot and when ESXi comes up it is not able to get a DHCP address or automatically assign itself an IP address

3) ESXi fails to load after a reboot with the error: PANIC: Error while reading file: -5

4) Keyboard stops responding when booting ESXi

5) You can't create a datastore on an array that is larger than 2 TB

6) After an install of ESXi, your NIC is listed as vmnic0 but in the DCUI has a status of disconnected

7) After an update, ESXi will not start VMs that are set to auto start

8) When you boot ESXi you get the errors "cannot setup ramdisk boot image is corrupted" and "intermediate name not a directory for /usr/bin" followed by a purple screen of death (PSOD)

9) When installing ESXi you get the error message "Unsupported BIOS setting (CPUID is limited). Disable 'limit CPUID value' or 'Support legacy/NT4 OS' in BIOS

10) Hostd.log is full of "Task Created : haTask-ha-host-vim.DiagnosticManager.browse-8375" messages

11) You create a USB flash drive to boot ESXi from and get the error: warning: bank 0 partition type invalid" for both bank 0 and 1

12) After you install your ESXi host you reboot and get I get " PANIC: Failed to find HD boot partition"

13) You use a custom oem.tgz to add support for a controller during the install, but it doesn't work after the install

14) var/log/messages has the error: Peer table full for sfcbd

15) Installing a patch fails with the error: "Unable to copy install files to host"

16) You try to add storage and get the error: "unable to read partition information"

17) When I try to add to add an existing VMFS datastore to my host, ESXi wants to format the storage

18) I've lost the root password to my ESXi host

19) I've expanded my storage array or LUN, how do I extend my VMFS datastore

20)When you try to create a new datastore, you get the error "Failed to Update Disk Partition Information"

21) During an install you have this error "Error not finding device to write"

22) You try to create a virtual disk and get the error: "The file is too big for the filesystem"

 

1) During an install of ESXi you encounter the errror "Unable to find a supported device to write the VMware ESX Server 3i 3.5.0 image to"

This error can have two meanings. First, that ESXi is not able to recognize the storage controller that you're trying to use for the install. Check the hardware list here to see if you have a controller ESXi will be able to recognize and load a driver for. If you have a controller from a vender that is not listed, then you may have to find a driver that will work with ESXi. Check this list to see if a 3rd party driver has been created. If your vender is listed, but the PCI id for your controller is not, then you can try adding support with a custom oem.tgz file. You can use the instructions here to find out the PCI id for your controller.
The second cause of this is that ESXi may be using an IDE driver for your controller. In this case, you should check the PCI id against the list mentioned to see what driver is listed for it. If you access the console during the install and login with the user root with no password, you can also run lspci -p and if you see an vmhba(x) device with a module of IDE, then ESXi has loaded an IDE driver for the controller. This is common with Intel ICH controllers. In this case, you can either use the IDE install method, change the BIOS setting for the controller to another mode (as this will change the PCI id of the controller and perhaps the driver used) or you can use a custom oem.tgz file to update simple.map to load another driver for the controller.

2) After an install of ESXi, you reboot and when ESXi comes up it is not able to get a DHCP address or automatically assign itself an IP address.

The address listed is 0.0.0.0 on the DCUI screen. If you press F2 (Customize System) and then select 'Configure Management Network', you get the below message Restore Network Factory Settings - This option will reset all your network configuration. to the factory defaults. If you select F11, the IP address listed is still 0.0.0.0. The most likely cause of this issue is that ESXi is not able to recognize your network controller. As with the above issue, check the hardware list to see if your NIC will be recognized. If you don't see it check the custom oem.tgz page for a driver. You can also try the following.
1) Access the console and run the command ls /mod. If you see a driver that will work with your NIC, run the command vmkload_mod /mod/<drivername>. For example, with a NIC that might work with the e1000e driver, run the command vmkload_mod /mod/e1000e.o. If the driver is able to load for the NIC, then you should be able to run esxcfg-nics -l and see the NIC in the list.
2) Disable any unneeded NICs or other components in the BIOS as they may be conflicting with your NIC and try the above steps again.
3) Run lspci -p to see if a driver has been loaded for the PCI id for your NIC. If you see an entry for the NIC with a vmnic(x) entry, but no module listed, it may be that the driver with ESXi does not work with that particular NIC. ESXi tends to use drivers that are a bit older so some newer NICs may not be supported yet.

3) ESXi fails to load after a reboot with the error: PANIC: Error while reading file: -5

This will typically indicate a corrupt file. In the case of a corrupt state.tgz, this is the file containing the configuration for your host. If you've taken a backup with vicfg-cfgbackup, then you should run a repair install with your ESXi install CD and then restore your backup file. A repair install will not overwrite your VMFS datastores so you won't loose your VMs (if you pick the repair option and not a new install). If your VMs are missing after restoring the backup config file, browse your datastore, find any VMX files and right click on them to add them back to inventory. If you had recently patched your host, you can press SHIFT+R during the initial ESXi boot screen to revert back to the prior firmware version. This will load the state.tgz file that was in use at the time of the last patch so any configuration changes made after that will be lost. VMs created after that can still be added as described above.
If you are getting this error while booting from the ESXi install CD, verify that your download is not corrupt.

Kernel: vmkerkel.gz .................................................................
Loading vmkernel.gz
Module: binmod.tgz
Loading binmod.tgz
Module: environ.tgz
Loading environ.tgz
Module: cim.tgz
Loading cim.tgz
Module: oem.tgz
Loading oem.tgz
Module: license.tgz
Loading license.tgz
Module: state.tgz
Loading state.tgz
PANIC: Error while reading file: -5, state.tgz
boot:

If you see the error PANIC: Error while reading file: -3 this may indicate a problem with the RAID controller that you're trying to install to.

4) Keyboard stops responding when booting ESXi

In some cases ESXi will seem to loose keyboard connectivity when booting. You can try to press ALT+F1 and ALT+F12 to see if you can see any error messages. If you don't get any response, try changing to another type of keyboard. If you're using a PS2 keyboard switch to a USB keyboard or vice versa.

5) You can't create a datastore on an array that is larger than 2 TB

When you create an array for ESXi to use that is over 2 TB is size, you find that ESXi will only be able to create a datastore that doesn't include the first 2 TB of storage. So a 2.2 TB array will only show 0.2 TB available for use as a datastore. ESX and ESXi have an array limit of 2.0 TB (2.0 TB - 512 B). With ESX, it will use the first 2 TB of the array and ignore the rest. With ESXi, it will ignore 2 TB portions and use the rest. So for a 2.5 TB array with ESXi, the datastore will be 0.5 TB. With a 5.0 TB array, the datastore will be 1.0 TB (5.0 - 2.0 - 2.0).
With most RAID controllers it is possible to create mulitple arrays on the same physical disks. So if you have 4 x 1.0 TB drives, you can create a 1.5 TB RAID 5 array with the 4 drivers and then create another 1.5 TB with the remaining space. ESXi will be able to create a datastore on each. In this case you could also create a datastore on the first array, and then use an extent to expand your datastore onto the 2nd array. With using extents you would end up with one datastore that is 3.0 TB in size. ESXi datastores can be up to 64 TB is size if you join 32 extents together. Generally one shoould avoid the use of extents as the lose of one extent will mean you would loose the VMs that have data on that extent. But in the example given above, both arrays would be on the same physical disks so the chance of loosing one array and not the other is very slim.

6) After an install of ESXi, your NIC is listed as vmnic0 but in the DCUI has a status of disconnected

If you access the console, the NIC is listed as down and with a speed of 0 Mbps if you run esxcfg-nics -l. At the console you can ping the IP of your ESXi host, but you can't access anything on your network. Try to disable ACPI in the BIOS of the host and see if the problem is resolved.

7) After an update, ESXi will not start VMs that are set to auto start

If you check the setting for auto start they are correctly set. If this is an ESXi host in stand alone mode, check to see if lockdown mode is enabled. If it is disable it. Lockdown mode does not allow for login with the root account, and that is the account used to start the VMs without vCenter server.

8) When you boot ESXi you get the errors "cannot setup ramdisk boot image is corrupted" and "intermediate name not a directory for /usr/bin" followed by a purple screen of death (PSOD)

It may be the case that one of the tgz files that ESXi is booting with is corrupt. If you have replaced oem.tgz there may be a problem with that file. Boot with a Linux live CD to replace oem.tgz or run a repair install to fix the system files. If you have just patched the host you can press SHIFT+r at the initial ESXi boot screen to boot with the previously installed firmware.

9) When installing ESXi you get the error message "Unsupported BIOS setting (CPUID is limited). Disable 'limit CPUID value' or 'Support legacy/NT4 OS' in BIOS

To get around this error during an install or when booting ESXi, press TAB key on the initial boot loader screen for ESXi and add nocheckCPUIDLimit as shown in the first image. This will get you through the install. When you first boot ESXi after the install, press SHIFT+o during the boot loader screen and add the option again as shown in the 2nd image. After you can connect with the VI client, go to Configuration \ Advanced Setting and uncheck the option VMkernel.Boot.checkCPUIDLimit.


10) Hostd.log is full of "Task Created : haTask-ha-host-vim.DiagnosticManager.browse-8375" messages

You find that tasks panel is full of "browse diagnostic manager" messages and hostd is full of "Task Created : haTask-ha-host-vim.DiagnosticManager.browse-8375" and "Task Completed : haTask-ha-host-vim.DiagnosticManager.browse-8375 messages". Edit the hosts file at /etc/hosts

11) You create a USB flash drive to boot ESXi from and get the error: warning: bank 0 partition type invalid" for both bank 0 and 1

You may get the below error messages. Ensure that the BIOS of your host has USB emulation set to fixed disk as shown here.
cannot get bank 1 parameters.
warning: bank 1 partition type invalid. Ignoring.
cannot get bank 2 parameters.
warning: bank 2 partition type invalid. Ignoring.
Panic: no usable banks found.

12) After you install your ESXi host you reboot and get I get " PANIC: Faield to find HD boot partition"

There can be a few factors that cause this. As discussed on this thread a BIOS downgrade was the solution, but you may also find that changing the SATA controller to AHCI mode correct this. You may also want to find out what PCI id your controller has and then post that info in the install forum. If you have used a custom oem.tgz to get the install to work, make sure you have updated both copies on the install CD. ESXi has a copy in the root of the CD for the install and another within a dd image in install.tgz (image.tgz for ESXi 4.0). There's a great Linux script here that will create an ISO for you and will update both oem.tgz files. Or you can boot with a Linux live CD as shown in this example and replace the file in the Hypervisor1 partition manually.

13) You use a custom oem.tgz to add support for a controller during the install, but it doesn't work after the install

The ESXi install CD has two copies of oem.tgz. See the above tip for a fix on this issue.

14) /var/log/messages has the error: Peer table full for sfcbd

You may also see the below errors.

UserThread: ###: Peer table full for sfcbd
World: vm #####: ####: WorldInit failed: trying to cleanup.
World: vm #####: ###: init fn user failed with: Out of resources!

To eliminate the error run the command /etc/init.d/sfcbd-watchdog stop. If the errors stop then you can go to Configuration \ Advanced Settings in the VI client and set Misc.CimEnabled to 0 and then restart the host. You may find that it is just sufficient to set Misc.CimOemProvidersEnabled or Misc.CimIHVProvidersEnabled to 0 instead of Misc.CimEnabled.

15) Installing a patch fails with the error: "Unable to copy install files to host"

If you look at the hostd.log file you may see the error: "Nfc_AuthdServerLoop saw error: File error -- Failed to write to the target file: There is no space left on the device". Check the /tmp folder to ensure that there is plenty of space. At the console you can run df -h as shown below. If the / mount is running low on space you may have to free up space. You can also free up space in /var/log and /var/log/vmware by deleting older log files.

16) You try to add storage and get the error: "unable to read partition information"

In some cases there may already be partition data on the LUN or array so you may have to create the VMFS datastore from the console. Below is an example of this procedure.

fdisk /dev/disks/vmhba32:0:0:0
Disk /dev/disks/vmhba32:0:0:0: 500.1 GB, 500107862016 bytes
   255 heads, 63 sectors/track, 60801 cylinders
   Units = cylinders of 16065 * 512 = 8225280 bytes
 Device Boot    Start       End    Blocks   Id  System
   /dev/disks/vmhba32:0:0:1             1     26109 209720478+  fb  VMFS
   /dev/disks/vmhba32:0:0:2         26110     40000 111579457+  fb  VMFS
Command (m for help): m
   Command action
   a       toggle a bootable flag
   b       edit bsd disklabel
   c       toggle the dos compatibility flag
   d       delete a partition
   l       list known partition types
   m       print this menu
   n       add a new partition
   o       create a new empty DOS partition table
   p       print the partition table
   q       quit without saving changes
   s       create a new empty Sun disklabel
   t       change a partition's system id
   u       change display/entry units
   v       verify the partition table
   w       write table to disk and exit
   x       extra functionality (experts only)
Command (m for help): n
   Command action
   e   extended
   p   primary partition (1-4)
   p
   Partition number (1-4): 3
   First cylinder (40001-60801, default 40001):
   Using default value 40001
   Last cylinder or +size or +sizeM or +sizeK (40001-60801, default 60801): 50000
Command (m for help): t
   Partition number (1-4): 3
   Hex code (type L to list codes): fb
   Changed system type of partition 3 to fb (VMFS)
Command (m for help): w
   The partition table has been altered!
Calling ioctl() to re-read partition table.
   Syncing disks.
   /proc/scsi/ata_piix # vmkfstools -C vmfs3 -S datastore4 vmhba32:0:0:3
   Creating vmfs3 file system on "vmhba32:0:0:3" with blockSize 1048576 and volume label "datastore4".
   Successfully created new volume: 49716c60-d3c6f150-ce9b-0015170da79a
 

17) When I try to add to add an existing VMFS datastore to my host, ESXi wants to format the storage

You might have this problem in several scenarios. First lets say you want to share a LUN between multiple host. On the first host, you will create a datastore with the Add Storage wizard on the Configuration \ Storage screen in the VI client. If you do that on the 2nd or any subsequent hosts, it will try to format the datastore which you don't want to do. In this case, first make sure that each ESXi host sees the storage with the same LUN id. If that's not the case, then this won't work. Then go to Configuration \ Storage Adapters and click on Rescan. If you have the LUN ids setup properly, then the storage should show up on the 2nd host.

In some cases you may be wanted to add the storage back to the same host or to another host. You won't be sharing the storage, but perhaps you've had to move the storage due to a controller or server failure. After you've added the storage, you should go to Configuration \ Storage Adapter and then check to see if the storage is listed. If it is, click on Rescan. If it doesn't appear as a datastore, then check /var/log/messages for errors about a snapshot. If you see this in the log file, you'll have to go to Configurations \ Advanced Settings and change either LVM.EnableResignature or LVM.DisallowSnapshotLun from the default settings. You'll then click on Rescan again. For more information see pages 117 to 119 in this document. The default for LVM.EnableResignature is 0 and for LVM.DisallowSnapshotLun is 1. If you are adding a snapshot or replica of a datastore back to the same servers, just change EnableResignature to 1 and the run a rescan. If you are adding the storage to different servers, then change DisallowSnapshotLun to 0 and run a rescan. After you rescan your datastore should be listed again and you'll want to change these settings back to the defaults. If you change the EnableResignature option, your datastore may end up with snap as part of the name as shown in this example.

If you're trying to reattach a datastore which consisted of extents, the process is a bit more complicated and may require some support from VMware.

18) I've lost the root password to my ESXi host

See the solution here.

19) I've expanded my storage array or LUN, how do I extend my VMFS datastore

See the solution here.

20) When you try to create a new datastore, you get the error "Failed to Update Disk Partition Information"

This problem can occur if there was a partition from another OS install on the disk that you're trying to add as a datastore. Boot with a Linux live CD or similar and either delete the partition or create a new one as discussed here. Then boot ESXi and try to create the datastore again.

21) During an install you have this error "Error not finding device to write"

If you access the console, you can see your storage. In some cases this error will be generated if the array you are trying to install to is larger than 2 TB. If that's the case, you need to reduce the size of the array. See here for more information.

22) You try to create a virtual disk and get the error: "The file is too big for the filesystem"

When you try to create a virtual disk that is larger than 256 GB, you get the above error even though there is plenty of free space available on the datastore.

     

 

 

 

 

INSERT YOUR COMMENT - IF YOU HAVE A QUESTION PLEASE USE THE FORUM


Name (required)

Web Site (optional)

Email address (required - will not be displayed)

Comment (required)

Please enter code

 

Copyright © 2011 - Dave Mishchenko