Thu Nov 5 17:17:35 CET 2020: Shut down virtual machines on ESXi host 172.16.0.8, FreeNAS server freenas Thu Nov 5 17:17:35 CET 2020: /root/scripts/esxi/esxi-stop-all-datastore-vms.sh root@172.16.0.8 datastore=[STORE_ISCSI] (max wait time=180s Full list of VM guests on this server: Vmid Name File Guest OS Version Annotation 12 Raspbian [Store_iscsi] Raspbian/Raspbian.vmx debian10_64Guest vmx-15 13 Ubuntu [Store_nvme] Ubuntu/Ubuntu.vmx ubuntu64Guest vmx-18 16 win10 [Store_iscsi] win10/win10.vmx windows9_64Guest vmx-18 17 CentOS [Store_iscsi] CentOS/CentOS.vmx centos8_64Guest vmx-15 18 Alpine [Store_iscsi] Alpine/Alpine.vmx other4xLinux64Guest vmx-15 19 Alpine2 [Store_iscsi] Alpine2/Alpine2.vmx other4xLinux64Guest vmx-15 23 win7 [Store_iscsi] win7/win7.vmx windows7_64Guest vmx-17 24 OPNsense2 [Store_nvmevo] OPNsense2/OPNsense2.vmx freebsd12_64Guest vmx-18 25 DappNode [Store_nvmevo] DappNode/DappNode.vmx debian10_64Guest vmx-18 3 FreeNAS [Store_nvme] FreeNAS/FreeNAS.vmx freebsd11_64Guest vmx-18 4 OPNsense [Store_nvme] OPNsense/OPNsense.vmx freebsd11_64Guest vmx-18 VM guests on datastore STORE_ISCSI: Found 0 virtual machine guests on 172.16.0.8 datastore STORE_ISCSI Total shut down: 0 Total powered down: 0 Thu Nov 5 17:17:37 CET 2020: /root/scripts/esxi/esxi-stop-all-datastore-vms.sh completed
I am 99% certain this problem is due to a change in the behavior of ESXi 7.0 from previous versions. I could be wrong, but in my experience, versions 5.5-6.7 always use uppercase names for datastores. For this reason, and in order to prevent errors, I therefore wrote the script to force datastore names to uppercase. Now, it seems, I will have to modify the script so that it no longer does this.
I did try commenting out of these few lines of code responsible for changing datastore name to uppercase but it didn't help...
You still need to assign the third command-line parameter to the esxidatastore variable, like this:
esxidatastore=$3
github.com
As a matter of curiosity but not only...I've updated the script on the GitHub repository:
GitHub - Spearfoot/utility-scripts-for-freenas-and-vmware-esxi: All-In-One utility scripts for FreeNAS and VMware ESXi
All-In-One utility scripts for FreeNAS and VMware ESXi - Spearfoot/utility-scripts-for-freenas-and-vmware-esxi
Is there any reason why all of these scripts are in separate files? Wouldn't have been better to have scripts in one file, divided in sections and toggled them on/off by some boolean variable?
ssh root@"${esxihost}" esxcli storage nfs41 remove -v Store_nfsssh root@"${esxihost}" esxcli storage nfs41 add -H 10.55.1.2 -s /mnt/tank/VM_nfs -v Store_nfs
I wrote them for use on my own systems and wasn't concerned about having multiple files at the time. You can always combine them into a single file, as you suggest, if that's what you prefer.As a matter of curiosity but not only...
As it is now, at least to me, it's like "Inception" movie, I need to go deeper and deeper to proper script and have other preconfigured as well. And since they are in separate files it's easy to loose a track. Don't get me wrong, I'm grateful for your efforts, it's just a suggestion.
I've never experimented with removing datastores as my only goal was to gracefully shut down any VMs stored on a particular all-in-one system's FreeNAS-based datastores. Of course you're welcome to modify the scripts to suit your needs.
Unfortunately NFS datastore must be removed prior to restarting its source/Freenas. Otherwise you'll end up with inaccessible datastore and at the end you have use cli to get it removed from esxi...
I've never had to remove NFS datastores. I use the esxi-rescan-datastores.sh script in my FreeNAS startup script. I provided an example of this in the GitHub repository.
And I also have been using esxi-rescan-datastores.sh and I keep using it (with added NFS commands)...
I haven't tested the combination of TrueNAS 12 under ESXi 7.0, so it may be that the datascore rescan - which has worked for all previous combinations of FreeNAS + ESXi - may no not work with that combination.
