SCSI error on VM
-
I am running a PFSense on a VM. I am getting the following error message on the console. UNMAP failed, switching to WRITE SAME(16) with UNMAP BIO_DELETE. Attached is the complete error message. There are multiple VM's on this host and it is running of an ICSI NAS. The drives on the NAS are fine as well as the other hosts. Any suggestions?
CapturFiles-202111318_171180.jpg -
What hypervisor? What emulated drive type?
Has this been running for some time and just started doing this or is this a new install?
Steve
-
@stephenw10 This has been running for a while now, I usually dont go into the console. I am running VMware 7.0 on the host. All I did was change the datastore on the PFSense box to another ISCSI share. Not sure if that caused it or not, but when I logged into the console I saw these errors. I rebooted and they are still coming up. The controller an LSI Logic Parallel controller which is the standard in VMware. Also Vmware tools is running on PFSense.
-
Mmm, it's not an error I'm familiar with. If you change the data store back does it stop?
Is it possible the errors were always present and you just didn't notice until now?
Steve
-
@stephenw10 It is totally possible that I have had these errors for a while. I just want to make sure that it is nothing serious that I should be concerned about.
-
Mmm, I would not want to see that on a production device. I did have an m.2 SATA drive in a test box that through errors similar to that continually and never actually failed. But I would not have relied on that.
Steve
-
Hi guys,
I am getting the same error. Brand new install of ESXi 7.0 U3b, brand new install of pfSense CE 2.5.2.
However I am using the SCSI Controller named "LSI Logic SAS" rather than the Parallel one, as this was the default one selected by ESXi when I chose a FreeBSD 12 64-bit based VM.
Any relevant info greatly appreciated.
Many thanks
Danny -
Actually the identical error? There are numerous CAM errors you might see there.
-
@stephenw10 Looks like it to me, here's mine:
I've literally only just installed ESXi and pfSense. This error was present during install of pfSense too, and on every boot of pfSense.
I'm guessing the last line is in relation to me installing the Open-VM package. I can't reboot yet to test though, deploying something. I'll update when I've rebooted whether it helped.
-
Seems similar to: https://forums.freebsd.org/threads/zfs-unmap-and-vmware-esxi-troubles.77648/
Something in VMFS6?
Steve
-
For me this issue was caused by thin provisioning on the virtual hard drive. I followed VMware's instructions to "inflate" the disk to thick provisioning and I stopped getting these errors.
-
T Topogigio referenced this topic on
-
Hi guys, this happend today in my 23.09.01 vm in Esx7.0u3. disk is thick zero. what happend?
-
Exactly the same error?
-
@stephenw10 yes
-
New VM or one you just upgraded?
-
@stephenw10 this is a VM updated two weeks ago to the last version, has been made two years ago
-
Hmm, so it ran fine for 2 weeks and then threw that error?
Anything different happen that might have triggered it?
-
Any time I've seen errors like that in a VM and the VM disk was on a NAS, it's been because the hypervisor lost contact with the NAS temporarily and that disk operation failed.
It could be a temporary network glitch, or it could be that the NAS was updating/restarting something at that moment as well.
I frequently can get away with rebooting my NFS server for updates and the VMs running off there don't notice as long as I'm not actively working on them. But it's all a matter of timing/luck.
-
I have the same - ben like this since 2.4 (now on 23.09.1) does not seem to affect anything
The pfSense is the only powered on VM on the HOST. (latest ESXI 7) host has local two SSD disks mapped 1:1 to two ESXI data storages and uses pfsense own software raid1 to mirror over those two.
-
Hmm, that's connected via an mpt(4) device. How is the raid configured in pfSense?
