Question

Unable to install Enterprise or CE

  • 16 February 2024
  • 1 reply
  • 264 views

Hi Next!

I've been trying to deploy Nutanix on three HPE DL380 G10 without success.

Hardware config is exactly the same on the 3 nodes, 2 x Platinum 8260Mm 1 TB RAM, 6 X 3.82 SSD drives, 4 port onboard 1 Gbps NIC (not used), 1 x Infiniband adapter (not used) and 2 x 2 port HP CN 1100R 10Gbps SFP+ cards. All firmware update with latest available PSP (september 2023). 

First tried Foundation, version 5.5 with installer package fraser-6.5.5.1 and with either ESX 7, ESX8 (HPE Cutomized ISOs) and AHV version integrated in the installer package. Foundation timeout after several retries, and checking iLO server console looks like the installer is not able to assign IP addresses to any network interface so foundation VM is not able to contact the server. If I try to manually assign IP addresses after foundation fails it returns an error indicating that eth5 to eth8 (10 GBps ports) file is not available. If I try to deploy using phoenix ISO (created from foundation VM / with ESX and also with AHV) it generates the infamous "Unable to match system information to layout module" that indicates, you shall consult KB7138. Checking KB7138 and there is no clear answer for the issue, is more dedicated to Nutanix hardware and for the rest of vendors it just says that you should make HPE replace the board because it shall report DX380 instead of DL380, not really an option here. Tried foundation with and without LACP and ports with and without VLAN. (switches are HPE FlexFabric 5700 10 Gbps, tried LACP aggregations and single trunks)

Tried CE and after waiting more than an hour for each server to boot (using phoenix via iLO) the system begins installation but close to 2043 just fails (seen different kind of errors with CE installer, must common is that it just can't image the hypervisor), looks like is not able to write the hypervisor to the drive. For CE I'm selecting the first RAID0 logical drive as hypervisor, the second RAID0 logical drive for CVM and the rest of the drives for data. Tried using other drives but the result is the same. Cleared all Smart Array config and re-created the Raid 0 arrays but is the same result. iLO reset, firmware reinstalls, same results. 

Please, please some help! It is supposed that DL380 G10 shall work…

Thanks!!!


This topic has been closed for comments

1 reply

Badge

Hi Jallamasc01,

I have experienced this on a few deployments where the hardware isnt at a supported firmware level and the interface cards are not recognised during the foundation process..

here are a few things to check:

  1. Firmware is at the latest level.
  2. iLO IPs are configure and accessable from the foundation instance.
  3. the Hypervisor / CVM Networks are accessable from the Foundation VM / Instance.
  4. Secure Boot is not enabled for the foundation process.

Hopefully this will assist..

 

Best Regards

John