So, I've been running a unRAID on a physical machine and I recently decided to move to VMware.  I bought a separate machine and got a stable ESXi environment on it.  Using PlopKExec, I got a fresh USB key to boot into an unRAID VM.  I bought a 9211-8i card and installed the IT firmware on it.  I installed that into my new machine, along with the drives from the old unRAID box.  On boot up of the ESXi server, I saw that it detected the drives.  I also swapped over my production USB key for the unRAID VM to boot from.  It boots up fine, and has the old IP address and settings, but it does not seem to see the hardware.  

 

How should I go about troubleshooting this?  

 

Thanks,

Paul

5 hours ago, Taige said:

but it does not seem to see the hardware

Can you explain this a bit more?

So, I've been running a unRAID on a physical machine and I recently decided to move to VMware.  I bought a separate machine and got a stable ESXi environment on it.  Using PlopKExec, I got a fresh USB key to boot into an unRAID VM.  I bought a 9211-8i card and installed the IT firmware on it.  I installed that into my new machine, along with the drives from the old unRAID box.  On boot up of the ESXi server, I saw that it detected the drives.  I also swapped over my production USB key for the unRAID VM to boot from.  It boots up fine, and has the old IP address and settings, but it does not seem to see the hardware.  
 
How should I go about troubleshooting this?  
 
Thanks,
Paul

Did you enable device passthrough for the 9211 card?

Sent from my ONEPLUS A3000 using Tapatalk

I started to reply to this last night...  As it turned out, I hadn't added the passthrough device to the VM.  After I did, it came up and started working, right before the ESXi server hung completely...  I went into the BIOS and tried disabling some settings, thinking perhaps there was a conflict of some sort.  I have one other slot I could move the card to, but I've not tried bringing it back up since the ESXI hung last night...

16 minutes ago, Taige said:

I started to reply to this last night...  As it turned out, I hadn't added the passthrough device to the VM.  After I did, it came up and started working, right before the ESXi server hung completely...  I went into the BIOS and tried disabling some settings, thinking perhaps there was a conflict of some sort.  I have one other slot I could move the card to, but I've not tried bringing it back up since the ESXI hung last night...

what  Motherboard do you have?

It's a HP ML30 Gen9 server...  So, it's got their built-in ILO4, their B140i RAID controller, etc.  

 

I'm actually not using the built-in SATA for anything, and not using the ILO portion of the NIC (though I'm using the NICs, of course).  I'm booting ESXi from a 32 GB microSD card (card reader built into motherboard).  

Edited October 17, 20178 yr by Taige

if you can switch B140i to AHCI, then you can use MB ports as simple SATA ports..

for me, ESXi is very stable for a long time.. so you need to understand what makes ESXi hang. i'm booting ESXi from USB..

I changed the BIOS from UEFI to Legacy BIOS, and I think that also disables some of the abilities of the B140i.  I have a very old Synology that is running iSCSI for the datastore.  I would like to have some sort of mirrored drive setup locally on the server, so I can get rid of the Synology.  Not sure if I can do that natively within ESXi or not.

 

Unfortunately, I'm not going to be able to test much more until tomorrow evening at the earliest.  This ESXi server also runs my firewall, so it needs to be up and running for the Internet to work...  I am working on an alternative to this, so I can take this server up and down without affecting the Internet, but I'm not quite there yet.

20 minutes ago, Taige said:

This ESXi server also runs my firewall

are you running this because of high speed VPN? i have cheep Mikrotik router which handle my 100Mbit up/down connection very well. for high speed VPN i'm using another Mikrotik - Cloud Host Router as VM inside ESXi server.

Well, it's a home lab environment for the firewall made by the company I work for. 

Ok, so some good news...  I got my second ESXi box up and got my firewalls working in HA mode.  I brought up the unRAID VM (on the other ESXi server) and had a few minor issues (it seemed to have lost my shares, which had happened before on physical hardware - rebooting seemed to fix it).

 

I have had a couple of backups go to it now without trouble, but I ran into a problem...  When I try to add a docker container, I'm getting this:

 

Error: open /var/lib/docker/tmp/GetImageBlob652775487: read-only file system

 

I'm running two 3 TB drives for storage with a 4 TB drive for parity and two 250 GB SSDs for cache.  These all connect to a 9211-8i with IT firmware, version P20.

 

It appears that the one docker container I was actively using (Plex) is stuck updating the EPG...  I imagine this is related.

Edited October 19, 20178 yr by Taige

Ok, I backed up my appdata, removed my cache drives, ran a pre-clear on them, added them back, formatted them, restored the backup, added a docker, and all looked good. Then I edited a file from my appdata directory (to set my DDclient up) and tried to restart the DDClient docker, only to get errors. Now, if I try to add another docker, I get this again:

 

Error: open /var/lib/docker/tmp/GetImageBlob019683831: read-only file system

I may have figured something out...  I used two different types of cables from the 9211-8i.  One of them came with another server.  Another one was a third party cable I ordered online.  One of the cache drives is attached with the cable from another server, and the other is the new third party cable.  

 

To see if it was related, I removed the cache drives, then put back a single drive, one that I think is attached with the original cable (from another server).  It's only been a few minutes, but so far everything is fine.  

 

Can a bad cable cause this sort of symptom?

 

Thanks,

Paul

More info...  Things seemed ok for a while, so I stopped the array, and tried adding the other disk back.  At that point, VMware crashed.  So, I'm thinking it's a cable or drive problem...  Anyone have any known good source for the breakout cables to connect to a 9211-8i?

11 hours ago, Taige said:

Anyone have any known good source for the breakout cables to connect to a 9211-8i?

You wanna connect 9211-8i to Sata HDDs? then you need Forward breakout cables.. According source, i have no idea... search the forums, here for example: https://forums.lime-technology.com/topic/27636-choosing-forward-breakout-cables/

 

It looks like it's having trouble even with only the one Cache drive (though not to the point that the system crashes, like it was doing)...  I'm having strange issues related to my cache drives, specifically my Plex docker...  

 

So, I've decided to remove this Cache drive as well to see if that will stabilize my docker situation...  

 

....  BUT...  Reading your sig made me think of another idea...  Move both of the SSDs to the ESXi server itself, and add them as Datastores, then create two virtual HDs as cache drives, one on each SSD.  That would give me the benefit of backup for cache, and the side benefit of having SSD speed storage space available for other VMs...