vmware Update manager errors

Today I wanted to upgrade an old ESX 3.5 host to the much newer ESX 4.1. A thing I’ve done through vmware Update Manager a couple of times over the last couple of days – but today showed a different side of VUM. It started out with that I couldn’t even get to scan the host for updates, that got solved with re-installing the vCenter agent on the host after quite a bit of troubleshooting. Then the next problem came, when I tried to remediate the upgrade to the host, it goes on for just a minute or so, and then fails and crashes the VUM service on the vCenter server… After another few hours of troubleshooting I found this in the VUM logdirectory

VUM log directory

‘Activation’ 5384 INFO] [activationValidator, 188] ManagedObjectNotFound. Leave Validate. Failed for vmodl.query.PropertyCollector.Filter.destroy on session private target: session[7BE206AA-16EA-4AD0-88F7-B8A0467D6800]95748830-9C71-4BBD-8488-B17934154A50 [2010-09-23 15:21:08:548 ‘Activation’ 5384 INFO] [activationValidator, 188] ManagedObjectNotFound. Leave Validate. Failed for vmodl.query.PropertyCollector.Filter.destroy on session private target: session[7BE206AA-16EA-4AD0-88F7-B8A0467D6800]FB7A2330-9122-43C6-91F0-443012184115 [2010-09-23 15:21:08:551 ‘Activation.trace’ 6252 DEBUG] [activationValidator, 1028] —————————————————— Invoking logout on integrity.SessionManager:Integrity.SessionMgr session B4CAF89A-E6CB-4721-ADB3-10250114A699 —————————————————— [2010-09-23 15:21:08:551 ‘Activation.trace’ 6252 DEBUG] [activationValidator, 1094] Throw vmodl.fault.SecurityError Result: (vmodl.fault.SecurityError) { dynamicType = <unset>, faultCause = (vmodl.MethodFault) null, msg = “”, } [2010-09-23 15:21:08:551 ‘Activation’ 6252 INFO] [activationValidator, 308] Leave Validate. No user logged in. Failed for integrity.SessionManager.logout on target: Integrity.SessionMgr


After searching on the error on vmware.com and just google.com – I came across this post: http://www.techhead.co.uk/vmware-vsphere-update-manager-there-are-errors-during-the-remediation-operation-and-inaccessible-vms which pointed me in the direction that it had to do with the storage, but not the same as the post mentioned, as there was not VMs on the host, and no orphaned VMs or anything, but anyway, I disabled the iSCSI adapter and removed the dynamic discory IP’s and rebooted the server, whereafter the server only could see its local storage. Then on the first try to remediate it worked and approx 20 minutes later I had a ESX 4.1 host up and running.