I noticed this with one of my CX array where consumed size is greater than the user capacity.
Agent Rev: 7.30.10 (2.48)
Prom Rev: 5.20.00
Here is the o/p with getlun
LOGICAL UNIT NUMBER 33
Name: Thin LUN 33
Current Owner: SP B
Default Owner: SP B
Allocation Owner: SP B
User Capacity (Blocks): 4194304
User Capacity (GBs): 2.000
Consumed Capacity (Blocks): 8389120
Consumed Capacity (GBs): 4.000
Pool Name: Thin Pool 0
Raid Type: r_5
Disk Type: SATAII
Here it can be noticed that User capacity is greater than the consumed capcaity.
That is a very small LUN. The difference is the metadata.
This is discussed in the 'Capacity overhead' section on page 68 of EMC CLARiiON Best Practices for Performance and Availablity: Release 30.0 Firmware update.
I have a question related to this.
From the documents, For a thin LUN 'consumed capacity' is the physical space used by the LUN and for a Thick LUN it is the host-visible capacity used by the LUN.
Does the 'consumed capacity' of a Pool LUN includes the RAID overhead, as it is the physical space. Also does it include the metadata overhead (metadata overhead = LUN size * 0.02 + 3) also?
Use capacity already excludes the RAID and mapping overhead.