Centera Pool and Drives

Hi Ashwi

You cluster total capacity of 10.9 TB in the empty cluster resulted in 9.1 TB available capacity. With mirroring in your system this would have been 4.5 TB usable capacity. If your data allows for single instancing it might have been more.

Now to your questions: The application has used 6 TB. As you do not provide the protected user data it is somewhere in the region of 7.6 TB (give and take a few GB for System Resources and Metadata). Your Centera has saved quite a bit with single instancing.

The pools in Centera are virtual pools and the quota is a soft quota.

1.: you can make your pool quota larger than the available capacity without problems as it's a virtual pool only.
2.: The pool capacity lists all the objects an application has written and the size of these objects all together. In an environment without single instancing your 6 TB would have used 12 TB protected userdata. If you use less then it thanks to single instancing. In terms of monitoring you need to monitor the available capacity. Experience shows that having less than 8 or 9 % available capacity is not a good idea. This is only 5 or 6 % from your current available capacity. Depending on your growth of protected userdata (or decrease in available capacity) you propably should start thinking about adding more nodes.
3: there is no way to map pools to nodes or drives. Pools are virtual pools and allow separating access and capacity reporting.

In case of additional questions do not hesitate to also contact me offline at holger.jakob@informatio.ch.

Best regards, Holger