Re: TRCASPBAL and STRASPBAL does not seem to work

[Philipp Rusch <Philipp.Rusch@xxxxxxxxxxxx>]

Hello Mark,

you can read it from the "model-number" of each disk under which
protection scheme it is running:

DD002 for example is a model "6607-072" which means a 6607 (= 4GB drive)
running in an 8 disk RAID set (that's the most sophisticated algorhythm
since you only loose about 10 % of disk space overall), that means, the
additional RAID info is spread over 8 disks and the rest of the disks
(up to the controller specific maximum of units) on that channel is
protected by this as well and does not need to store RAID infos at all.
The DD009 disk is running "alongside" this set (6607-070) without any
degradation in capacity at all, "model" -070 means it's protected, but
not part of the set.
"Model" -074 at last means that this is a set of four disks, taking
about 20 % of capacity for the additional info to protect the set.
This is obviously not that economic as the 8-disk algorhythm.

For performance reasons, I would recommend you move some of your disks
off the first controller (DD009 and DD015) add another 6607 disk (they
are really cheap if you take a used one) and add them to the second
2741 controller called DC02.
The most powerful controller you own is the 2748 called DC03, where
I would connect all the 8 GB disks, although you got two different
types of drives (6713 and 6717); I think they only differ in rotational
speed (7200 and 10k rpm). This is not ideal, but it allows you to form
an 8 disk array ot of the 8 gb disks + one additional 6713 (DD020).

Ok , before you start doing anything like above, you should have a closer
look at your limits (disk capacity), because removing and adding disks
takes large amounts of space.

Some investment in time and you'll improve your setup, ok, there's more
things to do, but this one is simple.

HTH, Philipp



"Mark A. Manske" schrieb:

> Thanks Philipp,
>
> The hardware config is really odd, here is the controller breakdown
>
> DC01             2741-001   Operational           Storage Controller
>   DD009          6607-070   Operational           Disk Unit
>   DD004          6607-072   Operational           Disk Unit
>   DD003          6607-072   Operational           Disk Unit
>   DD002          6607-072   Operational           Disk Unit
>   DD001          6607-072   Operational           Disk Unit
>   OPT01          6321-002   Operational           Optical Storage Unit
>   TAP03          6390-001   Operational           Tape Unit
>   DD008          6607-072   Operational           Disk Unit
>   DD007          6607-072   Operational           Disk Unit
>   DD006          6607-072   Operational           Disk Unit
>   DD005          6607-072   Operational           Disk Unit
>   DD015          6607-070   Operational           Disk Unit
>
> DC03             2748-001   Operational           Storage Controller
>   DD024          6717-074   Operational           Disk Unit
>   DD023          6717-074   Operational           Disk Unit
>   DD022          6717-074   Operational           Disk Unit
>   DD021          6717-074   Operational           Disk Unit
>   OPT03          6321-002   Operational           Optical Storage Unit
>   D01            283F-001   Operational           Device Services
>
> DC02             2741-001   Operational           Storage Controller
>   DD014          6607-070   Operational           Disk Unit
>   DD013          6607-074   Operational           Disk Unit
>   DD012          6607-074   Operational           Disk Unit
>   DD011          6607-074   Operational           Disk Unit
>   DD010          6607-074   Operational           Disk Unit
>   DD020          6713-070   Operational           Disk Unit
>   DD019          6713-074   Operational           Disk Unit
>   DD018          6713-074   Operational           Disk Unit
>   DD017          6713-074   Operational           Disk Unit
>   DD016          6713-074   Operational           Disk Unit
>
> And yes we are raid protected ;
> Sadly - this is beyond my understanding of which units are "used"
> and which are the raid packs - there are 2 other controllers, but
> they are for tape units.
>
> This is a "converted" 620 to a 820, so we still have a rack of the
> 620 left here for some disk units, and some disk units are in the
> new 820 rack, plus the 620 had a "side-cart" with disk packs too -
> This has been the "cost-conscious" way to upgrade - and it will not
> be changing anytime soon as long as the powers to be do not just
> hand over blank checks to us.
>
> Disk units getting hit heavy normally are 5, 6, 16, 18, 23, and 24
> sometimes at 40-75% while the rest are at 2-5%.
>
> The idea/gulp/thought of backing off data, deleting, balancing, and
> restoring is scary; I don't know if I could get that much dedicated
> time for the next two months... Sadly too, due to too many canned
> packages, and too many years of too many cooks in the kitchen, the
> data files are spread in 18 different libraries; getting better, but
> so much clean up with no "visible" ROI keeps the "mess" in place.
>
> It is something I could look at when time is more available closer to
> February.
>
> Sorry for such a big post - and thank you for any other insights, or
> suggestions
> that do not involve hardware upgrades.
>
> -Mark
>

--- SNIP a lot ---