Hi Team,
Please provide me your inputs, in finding a solution for these alarms, we are seeing multiple alarms on datastores "Thin-Provisioned volume capacity threshold exceeded" even making them to "Reset to Green", Datastores are not over-provisioned.
Unfortunately you forgot to supply the relevant numbers - run vmkfstools against the datastore ....
If you run out of free resources or if the sf-files cant be expanded any more I highly recommend to rebuild the datastore from scratch as soon as possible !!!
Ulli
Hi Continuum,
Thank you for your inputs, we have free space in all the datastores, and the existing datastores are not over provisioned.
Where we are unable to suspect any thing, created a ticket with VMware support and below are inputs they provided, please provide your inputs on these.
The "sdelete" utility was created by MICROSOFT and is supported by them. We do not provide support for "sdelete". For questions regarding SDELETE, you'll need to consult with Microsoft.
The sdelete utility is intended to unmap unused blocks within the Windows Guest OS, which vmware has NO control. In instances where the VMware unmap utility isn't successfully reclaiming disk space, it's usually because the Windows Guest asn't releasing the space to be reclaimed by the hypervisor.
The VMware VMFS unmap utility is used to unmap unused blocks ONLY within the VMFS datastore, not the guest. Therefore there are two separate steps involved since these Windows is layered on top of VMware: unmapping within the GUEST OS and unmapping within the VMWARE OS.
A third party article for Sdelete is provided along with a link to the download as a courtesy.. The VMware KB article with instructions is below.
Using SDelete and vmkfstools to Reclaim Thin VMDK Space (third party article)
https://vswitchzero.com/2018/02/19/using-sdelete-and-vmkfstools-to-reclaim-thin-vmdk-space/
Microsoft SDelete v2.04
https://docs.microsoft.com/en-us/sysinternals/downloads/sdelete
Using the esxcli storage vmfs unmap command to reclaim VMFS deleted blocks on thin-provisioned LUNs
https://kb.vmware.com/s/article/2057513
Also they recommened, Host patching / HardWare Firmware Update / etc.,
could you please let us know the impact with these alarms., VMware support only taking about "sdelete" from windows guests not others.
Wow - VMware was not able to supply the commandline for vmkfstools ? - that is scary !
run
vmkfstools -P -v10 datastorename
and provide the output
Hi
Please find the out put and provide me your suggestions.
LOL
Thin-Provisioned volume capacity threshold exceeded must be the dumbest error message by a datastore that only does thin provisioning ....
With humans that would be called
Attention deficit hyperactivity disorder (ADHD)
No further questions ....
Ulli
Seriously now ....
are you using a NFS-server based on win2012 ?
then I would translate the error message as "Is this a windows NFS-server ? - really ????"
are you using a NFS-server based on win2012 ?
we are not using any NFS-Server win2012 for "Datastores", using NFS - storage direction connections to vCenter hosts.
The screen provided is a datastore from one of the cluster.
From VMware point of view some types of NFS-servers are categorized as "obscure stuff that nobody would really use ..."
One example would be a win2012 NFS-server ....
Are you using NFS-servers that come from the obscure box .... ?
We have a NAS server, which is supplying the storage from windows servers to other windows guests, some purposes. How we can suspect that??