Hi Ilya, Can you confirm that you are using fiber channel VMFS type as your primary storage? I have tried enabling the storage.overprovisioning.factor in the global settings previously and it made no difference. The docs show that VMware storage overprovisioning is only supported on NFS and iSCSI, so I am curious if this is working for you.
Thanks, Chris On Thu, Oct 10, 2013 at 11:38 AM, Musayev, Ilya <imusa...@webmd.net> wrote: > Chris, > > Current solution is to enable storage.overprovisioning.factor in global > settings. Long term solution as per discussion with Edison, who maintains > storage framework, is to do real lookup of actual space. > > I'm in exact situation as you are, I've overprovisioned by factor of 20 > (maybe an over kill). I also rely on vSphere monitoring for space alerts. > > Regards > ilya > >> -----Original Message----- >> From: Chris Sciarrino [mailto:chris.sciarr...@gmail.com] >> Sent: Thursday, October 10, 2013 10:21 AM >> To: users@cloudstack.apache.org >> Subject: Cloudstack System Capacity - VMFS Storage >> >> Hi, >> >> I have the situation where I have deployed a fairly large amount instances of >> the same template. This has increased the amount of allocated primary >> storage used to over 2.5 TB . However due to the usage of linked clones I am >> really only using a couple of hundred GB. >> The primary storage type is a VMFS datastore so I do not believe I can >> overprovision it in cloudstack. The problem lies when trying to deploy more >> instances it is failing because the amount I have allocated is over the >> disable >> threshold even though I have plenty of actual storage left. >> >> Is there any way around this? Or any way to make cloudstack see the actual >> storage usage on the VMFS datastores? >> >> Thanks, >> >> Chris > >
