VMware Cloud Community
SelimAtmaca
Contributor
Contributor
Jump to solution

Inaccesible Objects on VSAN

First, I noticed two Horizon Desktop Pools provisioning were disabled. Then I saw ESX03 was not responding. Management Network Test were failing on ESX03. Rebooted ESX Host and these problems are gone. However  if I check ESX03 from VCenter Web Interface,  I see 12 inaccesible vms on ESX03.

screenshot1.PNG

I searched for a solution on the interner and read several articles and all RVC commands I tried so far  is telling me that I don't have inaccessible objects.

vsan.check_state ClusterName returns:

2020-03-10 23:03:30 +0300: Step 1: Check for inaccessible vSAN objects

Detected 0 objects to be inaccessible

vsan.check_state ClusterName -r returns

Did not find VMs for which VC/hostd/vmx are out of sync

vsan.purge_inaccessible_vswp_objects ClusterName returns:

2020-03-10 23:08:20 +0300: Collecting all inaccessible vSAN objects...

2020-03-10 23:08:20 +0300: Found 0 inaccessbile objects.

Do you think I should just remove them from inventory or should check something else?

Thanks

1 Solution

Accepted Solutions
TheBobkin
Champion
Champion
Jump to solution

Hello SelimAtmaca​,

Thanks for confirming.

In that case, there are no Inaccessible vSAN Objects and these are just leftover vSphere inventory references (likely as a result of the Objects being deleted without the VMs being unregistered), I would advise proceeding with removing the from inventory.

Bob

View solution in original post

4 Replies
TheBobkin
Champion
Champion
Jump to solution

Hello SelimAtmaca​,

A couple of points here:

- RVC doesn't appear to detect any Inaccessible Objects and thus these could just be vCenter-side cached references - if possible to reboot vCenter, please do this.#

Please run vsan.obj_status_report <pathToCluster> to confirm this (e.g. no Objects showing '1/3' or '0/3') or validate via the Health UI (Cluster > Monitor > vSAN > Health > Data), if this is Green then it is likely cached references.

- You mentioned this is a Horizon environment and you had some issues - potentially these are namespaces of VMs to failed to provision or had issues and were cleaned up or otherwise removed.

If this is the case and you are also not missing any important VMs from the environment (and/or have 0 Inaccessible Objects) then it may just be a case of removing these from vSphere inventory (will likely have to use the nested 'Infrastructure actions' option).

- Via RVC you can try using vsan.fix_renamed_vms <pathToVms> but note that this is unlikely to do anything if they are actually gone/inaccessible.

Bob

0 Kudos
SelimAtmaca
Contributor
Contributor
Jump to solution

Hi TheBobkin,

I rebooted VCSA and inaccessible objects are still here.

Here is the output of vsan.obj_status_report computers/VDI_CLUSTER/ command: (I don't know what 3/3 and 4/4 means but I guess it means something good)

2020-03-11 16:27:39 +0300: Querying all VMs on vSAN ...

2020-03-11 16:27:41 +0300: Querying all objects in the system from vdiesx04 ...

2020-03-11 16:27:42 +0300: Querying all disks in the system from vdiesx04 ...

2020-03-11 16:27:43 +0300: Querying all components in the system from vdiesx04 .                                                                                        ..

2020-03-11 16:27:44 +0300: Querying all object versions in the system ...

2020-03-11 16:27:47 +0300: Got all the info, computing table ...

Histogram of component health for non-orphaned objects

+-------------------------------------+------------------------------+

| Num Healthy Comps / Total Num Comps | Num objects with such status |

+-------------------------------------+------------------------------+

| 3/3 (OK)                            |  211                         |

| 4/4 (OK)                            |  1085                        |

+-------------------------------------+------------------------------+

Total non-orphans: 1296

Histogram of component health for possibly orphaned objects

+-------------------------------------+------------------------------+

| Num Healthy Comps / Total Num Comps | Num objects with such status |

+-------------------------------------+------------------------------+

+-------------------------------------+------------------------------+

Total orphans: 0

Total v1 objects: 0

Total v2 objects: 0

Total v2.5 objects: 0

Total v3 objects: 0

Total v5 objects: 1296

I checked Cluster > Monitor > vSAN > Health > and everything under that is green. (Passed)

These VMs are Horizon Instant Clone VMs and these VMs are deleted when user logs off. So There are no important VMs among them. Looking at the results of the commands you recommended, Should I just remove them from inventory or any other check required?

Thanks a lot

0 Kudos
TheBobkin
Champion
Champion
Jump to solution

Hello SelimAtmaca​,

Thanks for confirming.

In that case, there are no Inaccessible vSAN Objects and these are just leftover vSphere inventory references (likely as a result of the Objects being deleted without the VMs being unregistered), I would advise proceeding with removing the from inventory.

Bob

SelimAtmaca
Contributor
Contributor
Jump to solution

Thanks a lot Bob.

0 Kudos