EVA 4400 disk size available not correct after disk upgrade?

[Tom Raus]

Hi all,

I recently replaced 3x 146GB FC disks with 3x 300GB FC disks. Since I did this we have the following setup:

8x146GB + 8x300GB disks.
Our disk drive failure protection is set to single.
We have 2 enclosures, and the layout of those enclosures are exactly the same!
Controller hardware: HSV300
XCS version: 09001000
XCS build: CR10EAlep
Management firmware: 1.00.00.080211


and still I have the following situation which doesn't seem right to me.

Capacity: 2708.7GB
Occupancy: 2132.51GB
Available vRAID0: 399GB
Available vRAID5: 319.24GB

since we removed 3x146GB and replaced them by 3x300GB shouldn't we have at least an increase of 550GB in vRAID0? And we still had some room available before I did the replacements of the disks.

What I did notice however is that some of the disks have different firmware versions, we have the following firmware versions HP00 - HP01 - HP04

Is there anyone who can shed a light into this dark situation?

If you need more, or different information please let me know.

[Uwe Zessin]

The EVA always generates pairs of disk drives in a disk group to define the VRAID-1 members. If the system pairs a small and a large disk drive, only the capacity of the smaller disk drive can be used. It requires some deep poking to find out about the pairs - do a "ls disk full XML" from SSSU, attach it as a .TXT file and somebody can look at the details.

Different disk drive models can have different firmware levels to be current. In other words: the is no single level for all models.

Well, while we talk about firmware - it is *STRONGLY* recommended that the controller firmware is brought to a current level - this will also fix some bugs in the dsk drive enclosure I/O modules.

[Tom Raus]

Hi Uwe,

Thanks for your reply already, this sounds indeed highly probably that this is the case with us..

please find attached the requested output.
Any help will be appreciated as I'm in desperate need of that space :)

[Tom Raus]

now that you mention that the EVA groups the disks per 2.. enclosure 1 has 5x 300GB drives, and enclosure 2 has also 5x 300GB disks.. would it be wiser then to move one of the disks of enclosure 1 to enclosure 2 so that we have enclosure 1 with 4x 300GB disks and enclosure 2 with 6x 300GB disks?

[Uwe Zessin]

(I'm a bit short of time and have to take a look at the data later, sorry - feel free to remind me if I have not responded after a couple of days).

The EVA tries to split the redundancy information across as many disk enclosures as possible. If you collect disk size per enclosure, you could make the situation even worse. The problem is that the exact policies are not published and could even change - so there is almost no chance to influence the behaviour be placing disk drives manually.