First of all, here's my configuration:
I have an Intel NUC with a 256GB M.2 SSD inside, and an HGST 1TB spinning drive in it. It is running ESXi 6.0. installed on a 64GB USB thumb drive plugged into the back. I have a linux server VM, and a couple of windows server VM's installed on it. One of my windows server VM is about out of space, so I looked to see if I could increase the size of the virtual disk, but the datastore is full. I decided it was time to replace the HGST 1TB drive with a larger 2TB drive.
So this is what I did. I have a USB3 SATA dock with a 2TB drive in it. I plugged that into the front of the NUC and created a partition on it called 2TB-Temp, and then moved all of the directories from the internal HGST datastore to the new datastore on the external drive using the datastore browser. It took quite a while, but was successful.
I then shut down the server, removed the HGST drive, and then installed a new 2TB SSD inside, then plugged it in and powered it up. I created a new datastore on that new internal 2TB drive. So far, so good.
I then plugged the external dock with the 2TB-Temp drive in it back in. After plugging it back in, it now sees the drive under Configuration, Storage, Devices, but under Datastores, it doesn't show the 2TB-Temp drive now. I have been using the SSH shell into the ESX server to try and troubleshoot the issue, but can't figure out how to get the datastore to show up.
When I run the command:
esxcfg-scsidevs -m
I get this message for the drive in the USB dock:
VmFileSystem: SlowRefresh() failed: Cannot open volume: /vmfs/volumes/6275687b-e3e41838-3409-b8aeed793534. Cannot open volume: /vmfs/volumes/6275687b-e3e41838-3409-b8aeed793534mpx.vmhba39:C0:T0:L0:1 /vmfs/devices/disks/mpx.vmhba39:C0:T0:L0:1 6275687b-e3e41838-3409-b8aeed793534 0 2TB-Temp
t10.ATA_____CT2000MX500SSD1_________________________2137E5D2D6FD________:1 /vmfs/devices/disks/t10.ATA_____CT2000MX500SSD1_________________________2137E5D2D6FD________:1 627e97d6-a7bb3344-9ff4-b8aeed793534 0 Crucial SSD 2TB MX500
t10.ATA_____Samsung_SSD_850_EVO_M.2_250GB___________S24BNXAG901579B_____:1 /vmfs/devices/disks/t10.ATA_____Samsung_SSD_850_EVO_M.2_250GB___________S24BNXAG901579B_____:1 56811f06-aaf43310-d7d3-b8aeed793534 0 Samsung SSD 250GB
I've been searching through both the spiceworks KB and VMWare's KB to figure out what is causing this problem but to no avail. I did notice that the mpx.vmhba number is different now than when I originally formatted and created the VMFS partition on the external drive prior to moving the data to it. Is that perhaps the problem?
I've tried moving it to different USB ports on the front of the NUC to see if that would make a difference, but still can only see the device and not the datastore.
Any guidance or direction you could give me would be great.
ESXi 6.0 is no longer supported.
Personally I would put in your previous HDD, and either export the VMs as OVA or back them up , then swap your drives and import them
Have you tried a refresh on the datastores or a reboot of the host with the drive plugged in?
If you previously mounted the external USB as a datastore, did you cleanly unmount it or just pull it out?
Thanks for the reply Rod-IT.
I have rebooted the host multiple times with the drive plugged into both front USB ports. That's the only place it has ever been plugged in. I have also done the refresh and the scan for new devices multiple times as well. No change. I also never pulled the USB drive out while the server was up.
As far as putting the old drive back in and exporting the VM's, that's not an option either. When I used the datastore browser to move the data to the external datastore, it removed it from the internal datastore, so there's nothing on that drive to move. I even moved the scratch location to the external drive before shutting down and replacing the drive.
Hence my desperation. One of the windows servers is my exchange server, so all email is currently not available, and new email is being kicked back saying undeliverable now. It also has file shares on it with countless pictures and documents.
I would like to upgrade to a newer version of esxi, but thought first to upgrade the HD space, then upgrade to newer software.
The other thing I forgot to mention, was that if you do an ls -l of the /vmfs/volumes directory, you get this output:
When you look at the output in putty, it shows the first entry (2TB-Temp) pointing to a red text name for the volume, everything else is listed in blue. Does this indicate a broken link? Also, the error right after the ls -l command shows an input/output error for that volume. Hope this helps.
I hope this is only your home lab, especially if you have Exchange on a single HDD!
If you did not unmount the datastore, you may have to remove it manually from ESXi before you can re-add it.
As per your screenshot VMware is giving you an error on this, probably because it can't mount a volume when one with the same GUID already exists - at least to ESXi because you did not dismount it first.
Why did you use the move option and not the copy - I'm sure based on what you said this data is valuable, at least to you.
See if this link helps you, but for your sake I really hope you have a backup.
https://www.vcloudnotes.com/2016/07/unmounting-lun-or-detaching-datastore.html
Yeah, this is my home server. I worked in IT for over 25 years doing Windows, Linux, and AIX administration (with a little bit of ESXi at the end), but about 3 years ago, I burned out and walked away and have been doing hands on maintenance and repair work, so my IT skills have gotten a bit rusty. There's a lot of good information in that link. However, I get different results from the commands than what was shown in the link.
For example, when I run "esxcli storage vmfs extent list", I don't show naa id's:
[root@esxi1nuc:~] esxcli storage vmfs extent list
Volume Name VMFS UUID Extent Number Device Name Partition
--------------------- ----------------------------------- ------------- ------------------------------------------------------------------------ ---------
2TB-Temp 6275687b-e3e41838-3409-b8aeed793534 0 mpx.vmhba40:C0:T0:L0 1
Crucial SSD 2TB MX500 627e97d6-a7bb3344-9ff4-b8aeed793534 0 t10.ATA_____CT2000MX500SSD1_________________________2137E5D2D6FD________ 1
Samsung SSD 250GB 56811f06-aaf43310-d7d3-b8aeed793534 0 t10.ATA_____Samsung_SSD_850_EVO_M.2_250GB___________S24BNXAG901579B_____ 1
The drive I'm looking for is the 2TB-Temp, so I'm guessing my 'naa' number is mpx.vmhba40:C0:T0:L0
Am I understanding that correctly?
For the next steps, it says to unmount the datastore, but it doesn't show up in the Datastores list. If I proceed to the next step to detach the device, I get this error:
Call "HostStorageSystem.DetachScsiLun" for object "storageSystem" on ESXi "192.168.3.20" failed.
Not sure what to do about that.
When I try to list the permanently detached devices, I get nothing:
[root@esxi1nuc:~] esxcli storage core device detached list
[root@esxi1nuc:~]
Does that mean it's not detached?
When I try to list the filesystems, I get this error:
[root@esxi1nuc:~] esxcli storage filesystem list
Error getting data for filesystem on '/vmfs/volumes/6275687b-e3e41838-3409-b8aeed793534': Cannot open volume: /vmfs/volumes/6275687b-e3e41838-3409-b8aeed793534, skipping.
[root@esxi1nuc:~]
It doesn't even show me the filesystems for the internal drives.
I think I might still be seeing the old internal drive in the devices list -- or maybe it is the old ID of this same external USB drive from before shutting down and removing the internal drive, but not sure how to detach it since that menu option isn't there when I right click:
Maybe if we clean up the device list we might have better luck?
Thanks again for your patience with me.
Also, if I try to detach the device from the cli, I get this:
[root@esxi1nuc:/vmfs/volumes] esxcli storage core device set --state=off -d mpx.vmhba40:C0:T0:L0
Unable to set device status. Vmfs volume mounted on device with name mpx.vmhba40:C0:T0:L0
[root@esxi1nuc:/vmfs/volumes]
Does that mean I still have a datastore mounted somewhere on this drive? It doesn't show up in the list of datastores.
I'm so confused.
Your daily dose of tech news, in brief. Welcome to Friday! It has been a big week here as we launched Spiceworks News & Insights a few days ago. Do you know who else had their sights set high? Kenneth Arnold. On June 24, 1947, civilian pilot Ken...
I have a win 10 pro machine with 21H2.Running chrome 102.0.5005.115. I will open a bunch of tabs during the day (right now, 49). And including other things that are running, there's 80% of the 12GB of RAM in the box.I've noticed that when I go into ...
HI Spiceworld,I was reading some discussions around the community and I see that here it’s like we’re all part of a family, so I venture to open this discussion, hoping that you can help me/understand me or at least bring your experiences. I am a c...
Compulsion: 1: a very strong urge to do something He felt a compulsion to say something. 2: a force that makes someone do something She was acting under compulsion. 3: an act or the state of forcing an action They ...
I have a situation where one of the offices is looking to sublet for 1 day a week. They want one port to put their firewall/router on and drive their network from there. My first notion is to say no and require them to pull in their own connection. I want...
