Here is my humble CE lab housed in a Lack Rack.
3x HPE DL360p G8
2 x 6 core 2.6 ghz per node
64gb ram per node (more on the way)
1 x 500GB WD blue SSD per node
HP Aruba 2530 L2 switch
Lab looks great @Robwith2Bs - very clean setup.
how did you get the disks working I have an DL G7 and with the new 5.18 CE all SSDs are only recognized as HDDs. I think the problem is the P410i RAID Controller?
Did you already upgrade to 5.18 and if so, how did you get it to work?
Yeah the controller in the G7s will not support pass through mode. You will need a new HBA.
I have dl380 gen8 servers with 420i raid controllers and i cant seem to figure out how to get the hba pass through mode working on the raid controller. any help would be appreciated to point me in the correct direction.
I have dl380 gen8 servers with 420i raid controllers. I simply used the Provisioning tool and I create an array for every drive. Then it will set it up in Raid 0 and the drives are seen by Nutanix.
However, after getting everything installed and I login to the CVM it seems only one of the services is starting. So I am stuck at that point.
I am doing this with a dl380g8 also and can confirm redtech’s statement. I was having a similar issue with one of my SSDs and had to boot into the Provisioning tool using F5 at startup. I deleted all the logical disks and reassigned them, my problem SSD is now available and allowed me to install CE.
Once everything looks good, try the command “cluster start”
I tried that and all it does is say its trying to start all of the servers and repeats that message indefinitely.
I then tried on a Gen 9 and couldn’t even get it to boot. This CE is SOO finicky.
I think I’m giving up
I wanted to share the response I got from someone else running CE.“In my case it was because the cluster wasn’t able communicate over the internet to Nutanix to pull the license, make sure you can ping www.google.com (example) from the CVM. Making sure DNS and external internet access is possible…”I hope this helps give some direction to a possible cause. I’ll be working on getting my cluster going and can post an update if I get a resolution.
I think you are on to something. On the Host I can ping google.com. But in the CVM if I ping google.com I get “ping: google.com: Name or service not known”. I can ping my DNS servers IP address and I can ping 22.214.171.124 in the CVM though. So it does seem to have something to do with DNS resolution inside the CVM.
Sorry. took a break for vactaion. I did try messing with it a bit before but was unable to get the cluster started even after getting DNS assigned. I was able to find out it may require the HPE boot media to set the raid controller to HBA/Passthrough mode. Once I get more information, I can update, or possibly do a new thread.I’m not sure if this will be helpful but it does show going through setting the controller.https://www.youtube.com/watch?v=JuaezJd4C3I*note, mute the video. It plays music.
If you want to change HBA mode for the 420 or similar in a G8 Server , it’s much easier to download the ssaoffline-2.65-7.0.iso. ( https://support.hpe.com/hpesc/public/swd/detail?swItemId=MTX_c99ab52992ba4932a910178bc9 )For that download, you don’t need an HPE Account. It’s a bootable ISO for the Smart Storage Administrator Offline. With that version, you are able to do this setting in the UI (HBA-mode).
Thanks. I was trying with 3.10 and other iso mounted media for that matter and console would eject it before fully loading it. I will try this one on USB and see what happens.
Already have an account? Login
Enter your username or e-mail address. We'll send you an e-mail with instructions to reset your password.
Sorry, we're still checking this file's contents to make sure it's safe to download. Please try again in a few minutes.
Sorry, our virus scanner detected that this file isn't safe to download.