Solved

workaround for CVM storage listed as HDD

  • 22 March 2024
  • 8 replies
  • 217 views

Badge +1

We are trying out nutanix and in the installer, it asks me for CVM..

When i chose the 7mm that clearly is SSD it says its HDD and will not install on it..
Is there any workaround for installing cvm on this disk?

/pålt

7mm - Bay 0 LEN MTFDDAK240TDS-1AW1ZA  Normal 240 GB SATA SSD 2.5"”
icon

Best answer by ktelep 24 March 2024, 02:11

View original

This topic has been closed for comments

8 replies

Userlevel 4
Badge +5

Odd, can you show the output of the:

cat /sys/block/sdc/queue/rotational command?

Remember after setting it to 0, DO NOT REBOOT, just restart the installer.

Userlevel 4
Badge +5

How is the disk attached to the system?  What kind of host and what disk controller are you using? Are you using some form of RAID adapter or HBA?   Typically this issue crops up when the controller between the disk and the host can’t (or won’t) let the OS know that it is an SSD.

There is a workaround, but I’d recommend resolving whatever issue is causing it to show up as an HDD if possible.

Workaround is to first identify the sdX name associated with the device in the disk list of the installer, then hit Ctrl+C to break out of the installer and back to the command prompt.

Run:    cat /sys/block/sdX/queue/rotational
Rotational should be 0 for the SSD and 1 for HDD.

If the drive returns 1, the system believes there is no SSD

To fix this, set rotational to:
echo 0 > /sys/block/<device>/queue/rotational

Then restart the installer (do not reboot): 

./ce_installer && screen -r

 

Badge +1

Yes its in a raid..  The server was originally set up with esxi.. but we want to test out nutanix..
It might be some firmware updates.
I can do to check if it solves the issue 1st.. 
Thanks for the workaround answer

Badge +1
didnt work with firmwardate
 

 

Badge +1

got to installed now.. it boots on the USB Hypervisor.

 

Badge +1

Now i have another problem when creating the cluster. probably related to this:

nutanix@NTNX-30c53b0c-A-CVM:1x.x.x.x:~$ cluster --cluster_function_list="one                                              _node_cluster" --svm_ips=x.x.x.x create
2024-03-26 09:48:49,695Z INFO MainThread cluster:2943 Executing action create on                                               SVMs x.x.x.x
2024-03-26 09:48:49,704Z INFO MainThread service_utils.py:1615 The model type LO                                              GICAL VOLUME for SVM boot disk /dev/sdb is not in disk inventory, defaulting to                                               DAS-SATA
2024-03-26 09:48:49,704Z ERROR MainThread cluster:923 Cannot create one node bac                                              kup cluster as not enough boot ssds are present to facilitate backup on this clu                                              ster Available no of boot ssds is 1
2024-03-26 09:48:49,704Z ERROR MainThread cluster:3106 Operation failed
nutanix@NTNX-30c53b0c-A-CVM:x.x.x.x:~$
 

any more workaround that needs to be done?

this is the output from the hypervisor:
[root@NTNX-30c53b0c-A ~]# lsscsi
[0:0:0:0]    disk    Samsung  Type-C           1100  /dev/sdb
[15:0:0:0]   disk    ATA      ThinkSystem 7mm  00-0  /dev/sda
[17:0:0:0]   process Marvell  Console          1.01  -
[18:0:0:0]   enclosu 5350-8i  Smart Adapter    5.91  -
[18:1:0:0]   disk    5350-8i  LOGICAL VOLUME   5.91  /dev/sdc
[18:2:0:0]   storage 5350-8i                   5.91  -
[root@NTNX-30c53b0c-A ~]#
 

lsscsi from the cvm is:
nutanix@NTNX-30c53b0c-A-CVM:x.x.x.x:~$ lsscsi
[1:0:0:0]    cd/dvd  QEMU     QEMU DVD-ROM     2.5+  /dev/sr0
[2:0:0:0]    disk    ATA      ThinkSystem 7mm  2.5+  /dev/sda
[2:0:0:1]    disk    ATA      LOGICAL VOLUME   2.5+  /dev/sdb
nutanix@NTNX-30c53b0c-A-CVM:x.x.x.x:~$
 

Badge +1

so i did the same thing for the datadisk:
nutanix@NTNX-30c53b0c-A-CVM:x.x.x.x:~$ cat  /sys/block/sdb/queue/rotational
1
nutanix@NTNX-30c53b0c-A-CVM:x.x.x.x:~$
 

nutanix@NTNX-30c53b0c-A-CVM:x.x.x.x:~$ sudo echo 0 > /sys/block/sdb/queue/rotational
-bash: /sys/block/sdb/queue/rotational: Permission denied
 

But i can not write to it.

Badge +1

 “cluster --redundancy_factor=1 -s x.x.x.x create” worked for now.. At least it creates the cluster (without a backup cvm).. lets see how this goes..
Cluster created but everything down..
cs | grep -v UP is slowly giving less more down:
 

The state of the cluster: start
Lockdown mode: Disabled

        CVM: x.x.x.x Up, ZeusLeader
                               APLOS DOWN       []
                     PlacementSolver DOWN       []
                               Lazan DOWN       []
                             Polaris DOWN       []
                              Delphi DOWN       []
                            Security DOWN       []
                                Flow DOWN       []
                             Anduril DOWN       []
                               XTrim DOWN       []
                       ClusterHealth DOWN       []
 

and now all is up.. and i can https://x.x.x.x:2100 (but its; This page is deprecated for all cluster operations) main issue with disks solved