Betreff: HPE VMEssential stuck at 99%

[KevinAppasamy]

Starting VME Manager Services... ────────────────┐
│Progress: 99%

i installed the latest version of VMe found in HPE Software center

my netplan:

network:
version: 2
ethernets:
ens15f0:
addresses:
- 192.168.10.138/24
routes:
- to: default
via: 192.168.10.253

[BMuellerDSR]

Is it a nested installation on vmware / hyper-v? If so, make sure you expose the cpu virtualization extensions to the vm to enable nested virtualization.

[KevinAppasamy]

Hello, it is not nested. I’m trying to set up a single-node VM on Ubuntu 24.04 for testing purposes. I’m only using 1 node with local storage and want to use only 1 port. Is this possible?

I’ve connected my f0 port to the switch, but when I manually start the VM, it says I need to activate the “Management network” and create a “mgmt” bridge. I’ve assigned an IP to this bridge, and my f0 port is my SSH port, while f1 is for the bridge management. I’ve also given this bridge an IP, but it still doesn’t work.

Does the VM Manager need a separate IP from the bridge? The HPE VM Console says the VM is running (manually started), but I can’t ping it or access it via URL. My server does have access to 1.1.1.1.

[dya]

https://support.hpe.com/hpesc/public/docDisplay?docId=sd00006775en_us&page=GUID-B96D6475-9834-4E0F-A08B-CB97F532B786.html#ariaid-title7

Based on the description “Starting...”, I understood that the process was stuck at 99% or beyond on the following screen under “Manager Installation” at the above URL.

The exact cause is unclear at this time, but the issue of not progressing past 99% has occurred when the host cannot establish necessary communication with the manager for some reason.
For example, this could happen in the following situations:
・Incorrect IP address or similar
・Specifying an FQDN in the URL but failing to resolve the name

If the manager deployment completes successfully, tasks like creating the “mgmt” interface and assigning an IP address to it are also performed automatically during the deployment process.

If there is only one NIC, “ens15f0”, configure an IP address for ‘ens15f0’ during the OS installation. Then, during the manager deployment, specify “ens15f0” as the interface to use.

*Apologies if the machine translation is unclear.

[KevinAppasamy]

Thank you Dya,

i am on version 8.0.10, do you think there is an issue with that ? Is it mandatory to do a bond ?

My ens15f0 do have a free IP and i am using the VM IP as the URL.

[SergeMe]

This:

If the manager deployment completes successfully, tasks like creating the “mgmt” interface and assigning an IP address to it are also performed automatically during the deployment process.

Has not been my experience (also 8.0.10) when starting from a 'regular' Ubuntu 24.04 installation with all the required packages (KVM, Pacemaker, etc.) installed.

The mgmt OVS bridge does not get created automatically and if not present (and with required routing etc. in place if using a different subnet) the VM won't start just as the OP mentioned. 
Unfortunately the documentation leaves a lot to be desired about what is (supposed to be) happening in the background. 

[dya]

I hope this provides some insight into solving the problem.

・Confirmed on a virtual machine running Rocky Linux 9 KVM. 
・The version of HPE Morpheus VM Essentials is 8.0.10.
・The HVM host's IP address is 172.16.2.151. The manager's IP address is 172.16.2.150.
・Although the HVM OS installation ISO is recommended, I installed Ubuntu Server to proceed, thinking it would make the details easier to understand.
・Since the purpose this time is to share the sequence of steps, I did not bond the NICs.

Reference URL
https://support.hpe.com/hpesc/public/docDisplay?docId=sd00006775en_us&page=GUID-B96D6475-9834-4E0F-A08B-CB97F532B786.html

Points
・Installing the VM Essentials Console appears to install the necessary packages on the HVM host.
・On hosts deployed by the manager, an OVS bridge named “mgmt” is created and assigned an IP address.
・The manager VM connects to the OVS bridge “mgmt”.

Below, I checked what I did and the status of the network and package installation in several places.

-----------------------

1. Install Ubuntu Server 24.04
2. apt update and apt upgrade
3. Verify the current status

・Network
root@test-hvm:~# netplan get
network:
version: 2
ethernets:
enp1s0:
addresses:
- “172.16.2.151/24”
nameservers:
addresses:
- 172.16.2.8
- 172.16.2.9
search:
- home.example
routes:
- to: “default”
via: “172.16.2.254”
root@test-hvm:~#

root@test-hvm:~# netplan status
Online state: online
DNS Addresses: 127.0.0.53 (stub)
DNS Search: home.example

● 1: lo ethernet UNKNOWN/UP (unmanaged)
MAC Address: 00:00:00:00:00:00
Addresses: 127.0.0.1/8
::1/128

● 2: enp1s0 ethernet UP (networkd: enp1s0)
MAC Address: 52:54:00:5a:a6:4f (Red Hat, Inc.)
Addresses: 172.16.2.151/24
fe80::5054:ff:fe5a:a64f/64 (link)
DNS Addresses: 172.16.2.8
172.16.2.9
DNS Search: home.example
Routes: default via 172.16.2.254 (static)
172.16.2.0/24 from 172.16.2.151 (link)
fe80::/64 metric 256
root@test-hvm:~#

・Installed packages
root@test-hvm:~# apt list --installed | egrep -i “Libvirt|pacemaker|ceph”

WARNING: apt does not have a stable CLI interface. Use with caution in scripts.

root@test-hvm:~#

4. Install VM Essentials Console
5. Verify current status

・Network
root@test-hvm:~# netplan get
network:
version: 2
ethernets:
enp1s0:
addresses:
- “172.16.2.151/24”
nameservers:
addresses:
- 172.16.2.8
- 172.16.2.9
search:
- home.example
routes:
- to: “default”
via: “172.16.2.254”
root@test-hvm:~#

root@test-hvm:~# netplan status --all
Online state: online
DNS Addresses: 127.0.0.53 (stub)
DNS Search: home.example

● 1: lo ethernet UNKNOWN/UP (unmanaged)
MAC Address: 00:00:00:00:00:00
Addresses: 127.0.0.1/8
::1/128

● 2: enp1s0 ethernet UP (networkd: enp1s0)
MAC Address: 52:54:00:5a:a6:4f (Red Hat, Inc.)
Addresses: 172.16.2.151/24
fe80::5054:ff:fe5a:a64f/64 (link)
DNS Addresses: 172.16.2.8
172.16.2.9
DNS Search: home.example
Routes: default via 172.16.2.254 (static)
172.16.2.0/24 from 172.16.2.151 (link)
fe80::/64 metric 256

● 3: virbr0 bridge DOWN/UP (unmanaged)
MAC Address: 52:54:00:b7:91:a5
Addresses: 192.168.122.1/24
Routes: 192.168.122.0/24 from 192.168.122.1 (link)
root@test-hvm:~#

・Installation status of several packages (some results omitted due to length)

root@test-hvm:~# apt list --installed | egrep -i “Libvirt|pacemaker|ceph”

WARNING: apt does not have a stable CLI interface. Use with caution in scripts.

ceph-base/noble-updates,now 19.2.1-0ubuntu0.24.04.2 amd64 [installed,automatic]
libvirt0/noble-updates,now 10.0.0-2ubuntu8.9 amd64 [installed,automatic]
pacemaker/noble,now 2.1.6-5ubuntu2 amd64 [installed,automatic]
root@test-hvm:~#

★All packages required for HVM hosting appear to be installed on the host.

6. Installing the Manager

7. Current Status Check
・Network
root@test-hvm:~# netplan get
network:
version: 2
ethernets:
enp1s0: {}
bridges:
mgmt:
addresses:
- “172.16.2.151/24”
nameservers:
addresses:
- 172.16.2.8
- 172.16.2.9
search:
- home.example
interfaces:
- enp1s0
routes:
- to: “default”
via: “172.16.2.254”
openvswitch: {}
root@test-hvm:~#

root@test-hvm:~# netplan status --all
Unknown device type: openvswitch
Unknown device type: openvswitch
Unknown device type: openvswitch
Unknown device type: openvswitch
Unknown device type: openvswitch
Unknown device type: openvswitch
Unknown device type: openvswitch
Unknown device type: openvswitch
Unknown device type: openvswitch
Unknown device type: openvswitch
Online state: offline
DNS Addresses: 127.0.0.53 (stub)
DNS Search: home.example

● 1: lo ethernet UNKNOWN/UP (unmanaged)
MAC Address: 00:00:00:00:00:00
Addresses: 127.0.0.1/8
::1/128

● 2: enp1s0 ethernet UP (networkd: enp1s0)
MAC Address: 52:54:00:5a:a6:4f (Red Hat, Inc.)

● 3: virbr0 bridge DOWN/UP (unmanaged)
MAC Address: 52:54:00:b7:91:a5
Addresses: 192.168.122.1/24
Routes: 192.168.122.0/24 from 192.168.122.1 (link)

● 4: ovs-system other DOWN (unmanaged)
MAC Address: 56:63:d3:67:0f:2d

● 5: mgmt other UNKNOWN/UP (networkd: mgmt)
MAC Address: 56:8c:c5:0e:ad:40
Addresses: 172.16.2.151/24
fe80::548c:c5ff:fe0e:ad40/64 (link)
DNS Addresses: 172.16.2.8
172.16.2.9
DNS Search: home.example
Routes: default via 172.16.2.254 (static)
172.16.2.0/24 from 172.16.2.151 (link)
fe80::/64 metric 256

● 6: vnet0 tunnel/tun UNKNOWN/UP (unmanaged)
MAC Address: fe:54:00:67:0d:cb
Addresses: fe80::fc54:ff:fe67:dcb/64 (link)
Routes: fe80::/64 metric 256
root@test-hvm:~#

8. The manager VM is connected to the OVS bridge “mgmt”.

root@test-hvm:~# virsh list
Id Name State
------------------------------
1 test-manager running

root@test-hvm:~#
root@test-hvm:~# virsh domiflist test-manager
Interface Type Source Model MAC
---------------------------------------------------------------
vnet0 bridge Management virtio 52:54:00:67:0d:cb

root@test-hvm:~#
root@test-hvm:~# ovs-vsctl show
ac2e21e3-aa8e-449d-800d-5487a309a9ea
Bridge mgmt
fail_mode: standalone
Port vnet0
Interface vnet0
Port mgmt
Interface mgmt
type: internal
Port enp1s0
Interface enp1s0
ovs_version: "3.3.4"
root@test-hvm:~#

[dya]

I've made a few minor corrections to the above.
Sorry for the length.

[SergeMe]

No problem and thanks for taking the time to go through this!
I should have mentioned that my Ubuntu Server 24.04 is in fact an Azure VM and I deployed 2 of them to create a cluster. I suspect that is probably a complicating factor.
When I get a chance I will try what you did and simply use a flat network (for host and VMs) to see if the mgmt OVS bridge gets created. I think this could create some interesting routing dilemmas when trying to do this with 2 hosts and a flat network (with 1 subnet) with KVM hosts and VMs running on the same subnet. though

[dya]

Thank you for your confirmation.

The previous example was provided solely to illustrate that the Manager VM boots successfully  with a single NIC.

In practice, when I perform verification, I sometimes place the Manager VM and the virtual machine on the same VLAN, and sometimes on different VLANs.

I hope it works out for you.

 (Reference) Network Considerations
https://support.hpe.com/hpesc/public/docDisplay?docId=sd00006775en_us&page=GUID-8043BF83-F3FA-4A33-A7DE-C304B4F1BEEF.html
※Looking at it now, the content has changed significantly. I was surprised.