VCSA upgrade “error: vimService creation failed: Error:”

There has been a lot of issues involving the upgrade UI for the VCSA migration and upgrade utility. I met it at the “Connect to host appliance stage” Here are some things that I tried to resolve the issue with no success.

  1. Check DNS Forward and reverse
  2. Disable anti-virus on all applicable machines
  3. Reset the PSC passwords
  4. Reboot the PSC’s

The solution to this problem that I found was the build numbers I was trying to implement. Here is my example I will lay out for you.

Say you have a VCSA and an External PSC. Both are on the same version of code.

  • The VCSA and PSC are now on version 6.0 U3f 8874690 release date of 2018-06-28
  • My target version for the VCSA and PSC is  6.5 U1g 8024368 release date 2018-03-20

The problem:

The problem here that even though it is a newer version of VCenter it is an older version of code. So even though you are going to 6.0 to 6.5 the Upgrade utility will not allow you to upgrade it will give you this vague and unhelpful error.


Check your build numbers and always make sure the code you are upgrading to  matches the current version or is newer than the current version of code.

VXRail Hardware failures? Not so fast.

Recently had a VXRail upgrade and then noticed hardware failures on equipment we were not using. Worked with it through the DELL/EMC support cycles and we came to two conclusions. Update to the latest version of VXRail or reset the BMC on the troubled node. Because the hardware was not the issue. These are the steps they provided me and it worked.

  •  SSH to the VXRail node
  • Run the following command on the host
/opt/vxrail/tools/ipmitool mc reset cold
  • Once that command completed took about 3 minutes for me. Log into the VXRail Manager via SSH use the default “mystic” username and password.3
  • Restart the VMWare Marvin service and the Jars service. (Note you may need to su to root)
systemctl restart vmware-marvin
systemctl restart runjars

Once VXRail manager is back up log in and check the console to make sure the hardware errors are gone.


VMworld VBrownBag presentation 2018

This is my Presentation from VMWorld 2018 it was a tough topic and really interesting to try and explain to a crowd of people. One of the biggest problems I had was trying to overcome the fear that not everyone had the same experience as i had. But in the end it did not matter. This was my experience with priorities. I have a lot to work on for my public speaking skills but I think it has gotten better!

“Trust Relationship Between this workstation and domain failed”

This is a pretty annoying thing that happens when coming back from an old snapshot. Usually over a week. You roll back a snap and then find out you cannot log in with domain accounts. This sucks and makes the roll back more annoying than it has to be. Fortunately the fix is easy. As long as you have the local admin password handy or a cached domain credentials.

Steps below for local admin

  1. Log into windows box with local admin password.
  2. Open Up power shell as an administrator
  3. Type the following commands one after another.
     $credential = Get-Credential DomainCredentials
    Reset-ComputerMachinePassword -Server DomainController -Credentials $credentials
  4. Reboot the machine and log in.

Steps Below for cached credentials

  1. Log into VCenter server where VM is located
  2. Edit settings on VM
  3. Disconnect all network cards on the VM (Do not drop the NIC’s just disconnect)
  4. Log in with cached domain credentials. ( If this doesn’t work and you don’t have the local admin you are SOL)
  5. Open Up power shell as an administrator
  6. Type the following commands one after another.
     $credential = Get-Credential DomainCredentials
    Reset-ComputerMachinePassword -Server DomainController -Credentials $credentials
  7. Reboot the machine and log in

Vsphere Content Library: RuntimeFault.Summary cannot delete content library item

When trying to delete an image from content library and you receive an error of:


“Cannot Delete File [Datastore] /PathtoContentLibraryVMDK”

Solution I found:

  1. Log into VCenter Server
  2. Select Content Library
  3. Right click content library item
  4. Rename item
  5. Delete item

I was trying to update some images i had on my home lab and i ran into the issue and issue when trying to delete some old templates from the library. I tried multiple times to delete the template i had stored only to receive the same error.

I investigated and found that the path from the error had no VMDK in it from the content library. That being said I had two options. Delete the folder in the datastore, which could corrupt the content library or rename the library item and try to delete again.

So when I renamed the content library item and deleted the item and it worked like a charm. I could not find a KB article for this particular issue or at least one that was relevant to content library but this worked for me.


EMC SourceOne – Exchange 2016 “system call failed”

“System call failed.  Unknown error”

“CoExExchProvider::LogonDS: Cannot log into mailbox”

If after you migrated the SourceOne service account and journal mailbox to Exchange 2016, and your jobs keep failing you may have a similar issue to me.

During the process of migrating the last two mailboxes (Both SourceOne) to the new environment I was getting permission errors throw in the logs while attempting to run the journaling jobs. This was because SourceOne DOES NOT SUPPORT HTTPS. 

In order to get SourceOne working properly on exchange 2016 if you are forcing HTTPS you need to enable RPC over http using the KB articles below.

Each step must be done on each worker node after migration has completed:

  1. Log into worker node using SourceOne Service account
  2. Control Panel > Mail (32-Bit) > Delete old mail profile > Create new profile (Make sure to uncheck Cached exchange mode)
  3. Regedit.exe navigate to “HKEY_CURRENT_USER\Software\Microsoft\Exchange”
  4. Create new DWORD “MapiHttpDisabled”
  5. Set value to “1”
  6. Restart Outlook client

There is an KB from Microsoft that shows you how to remove this setting, so i did not add it because it may have been confusing. But the KB is 2937684.


EMC SourceOne 7.2 SP6- Cannot connect to journal mailbox in a mixed Microsoft Exchange scenario (2010 and 2016)

So this has bitten me for the past couple of weeks we would try to fail over our last two mailboxes from exchange 2010 to exchange 2016. One Journal and One Service account. Come to find out (No thanks to EMC SourceOne install documentation) That the service account needs to be the last mailbox to move.

Take a look at this bug report:

If you are going to run in a mixed environment for a long time follow the steps below in the article and always make sure you migrate your SourceOne Service account last!

With 2010/2016 co-existence, there are certain requirements for it to work.    
    These requirements are:-    
    1) Service Account mailbox needs to be always on the lower exchange version.It should be the last mailbox to move to 2016. Exchange 2010 should be configured with RPC over https.    
    2) Exchange 2016 should have RPC over https enabled. By default Exchange 2016 has MAPI over https which is not supported by SourceOne. Customer will have to enable RPC over https and configure it correctly    
    3) All ES1 servers should have only Outlook 2010 SP2 + KB2878254. No other HF or update. Updates for Outlook should be disabled.    
    4) Profile needs to be configured on Outlook to have Outlook using RPC over https with the proxy server pointing to the 2016 server. For these changes you will need to configure AutoDiscovery.    
    5) Set the correct permissions mailbox on 2016    
    Without these 5 criteria, they will not be able to communicate between 2 versions of exchange.