Dell Perc 6/i (RAID5) larger drives

We have a Dell Poweredge Server running SBS 2008 with a Perc 6/i controller and 3x 250GB HDDs. One of the drives on a RAID5 set has degraded so will need to be replaced. I thought that as the server was getting low on space it might be an idea to increase the size of the drives while I'm at it.

On other RAID controllers I've used, it's been quite straightforward but time consuming to simply replace the drives one by one letting each newly inserted drive rebuild and then after all drives have been rebuilt, use EaseUS Partition Manager to extend the partitions.

Is this possible with Perc 6/i and if so are there any gotchas I should be aware of?

I'm aware of the 2GB limit of Perc6 but does that apply to the final virtual disk size or the individual disks?

Happy to use an alternative partition manager if EaseUS is a no go. Will the Windows Disk Management 'Extend Volume' even work?

The server is well out of warranty and don't want to give the job to the overexpensive Dell engineers for something that in normal situations is quite straightforward.
Who is Participating?
PowerEdgeTechIT ConsultantCommented:
"Is this possible with Perc 6/i and if so are there any gotchas I should be aware of?"

No.  The only thing you will be able to do with the newly added space of replacing current disks with larger ones is create a new virtual disk (RAID array) across the disks.  This will appear to Windows as a new and separate "disk", so it cannot be merged with existing partitions (if your goal is to join or extend current partitions).  The PERC supports online expansion, but only by adding disks to it  ... for example, you could add one or two (or more) 250GB disks, then convert your 3-disk RAID 5 to a 5-disk RAID 5, THEN expand the partition sizes.

"I'm aware of the 2GB limit of Perc6 but does that apply to the final virtual disk size or the individual disks?"

The PERC is limited to 2TB PHYSICAL DISKS (not logical/virtual disks).
It is probably simpler and faster just to take an image-based backup while the server isn't connected to the LAN, then replace all the disks with the larger ones, build your new array and restore the backup. Many of those backup tools allow you to do the restore to larger partitions, so you wouldn't need to resize your partitions manually. When done, just connect the LAN again.

The extend volume works provided the unallocated space comes right after the partition you want to extend.

Paragon has several tools for backup or also partition changing:

You could also use the CloneZilla to image, and the Partition Editor on the PartedMagic LiveCD, but it'll be a little more difficult to use:
Actually the PERC6 does support some disks > 2TB.  First the chipset has no issues with SAS drives > 2TB.  The problem is specifically 512-byte sectored SATA disks.  But some 4K-sectored  SATA disks have been qualified.

Hitachi has claimed to qualify their 4TB SATA drives with the PERC 6/i :$file/US7K4000CompatGuide_ver1.0.pdf

As for a swap-disk-upgrade.  Don't do it.  You have risk of data loss.  Never, ever, ever degrade an array on purpose to update.  All it takes is one bad block and you are screwed.  You also can't just go back because once you start the rebuild you can't force the other drive back in, so you have to do a backup of what you can and then blow the whole raid away, and rebuild with the data loss.

It takes less time on the clock, and is safer to do it the right way.  If you don't have enough free space to back up, then i suggest you spend time getting a backup going first. You then have the perfect opportunity to try to do a bare metal restore/test a recovery when you migrate to new disks.

Just be sure to shut the system down and yank all the low-capacity drives before attempting the restore  That way if you have to go back, the disks *think* they are still active so the controller will accept them if your restore fails and then you can try another backup plan once you correct the problem that prevented a restore from working on replacement drives.
Free Tool: ZipGrep

ZipGrep is a utility that can list and search zip (.war, .ear, .jar, etc) archives for text patterns, without the need to extract the archive's contents.

One of a set of tools we're offering as a way to say thank you for being a part of the community.

PowerEdgeTechIT ConsultantCommented:
"Actually the PERC6 does support some disks > 2TB."

I'd have to see it (in action) to believe it.  My experience and docs/briefs I have I have seen from Dell indicates that the PERC 7 (H200/H700/H800) are the first controllers with the capability of using larger than 2TB disks.  You have confirmed the 2TB limit on PERC 6's before ... has something changed or is this a new discovery (not just an HCL)?
The problem is specific to the SATA bridge code, and is not a limit to the SAS.  The problem is with the chipset they (LSI) uses, and only manifests itself with SATA due to the bridge translation of ATA to SCSI commands.

I've got several controllers here hooked up to 3TB SAS drives and they all work fine.  One is even running Solaris.  The reason there is a perception out there that such a limit exists is because 99% of the people who try it use SATA disks.  [Latest firmware/drivers required of course]

The PERC6 is fundamentally a different controller than the PERC 7 which certainly has no issues.

The 2nd difference is that the 2TB SATA limit can be exceeded if you are using AF disks.  When I confirmed before the user had 512-byte sectoring so it was moot.

I have NOT tried 4KB sectoring on a PERC6.  Maybe this was a special for Dell, maybe not. As a developer I can say that the latest development source code is compatible with 4KB sectoring and I have no reason to believe it won't work.

What dell chooses to qualify is their business. I am saying that Hitachi claims they qualified 4KB SATA on that controller.  I am saying I know for a fact that the 2.09TB constraint is limited to SATA. I am also saying I know for a fact that the firmware in PERC6 uses 16-byte SCSI commands when necessary, so you can have LUNs in the petabyte range if you can find SAS disks that big on the PERC6.
PowerEdgeTechIT ConsultantCommented:
Guess I'll have to watch for an opportunity to test this myself.  (Not that I don't believe your logic and experience :), it's just that in order for me to tell people a controller can do something that ALL documentation indicates it cannot, I'd have to see it.)
I expect nothing less, PowerEdgeTech...  But just consider that the Documentation you have access to has more to do with what Dell wants to sell and support than what a product can do.

I did just double check to make sure I did have the 1068E running 3TB SAS drives, here are details so you know the specifics of what I use that does work.
Port #0. mpt0              RAID SAS1068E B1    MPT 105   Firmware (1.30.10)   IOC 0
x86 BIOS image's version is MPTBIOS- (2009.11.12)
EFI BIOS image's version is

Interrupt Coalescing:           Enabled, timeout is 16 us, depth is 4
SAS1068E's links are 3.0 G, 3.0 G, 3.0 G, 3.0 G, 3.0 G, 3.0 G, 3.0 G, 3.0 G

 Bus Tgt L  Vendor   Product          Rev      SASAddress     PhyNum  Enclosure         Bay   Serial               OSName
  0   10 0  SEAGATE  ST33000650SS     0002  5000c5003425b739     3    500605b00027d7b0    3   Z2905CL0             /dev/rdsk/c2t10d0s2  

Open in new window

The SEAGATE drive above is a 3TB SAS and I verified long ago that it reports full capacity.

I dropped an email to one of my contacts to clarify AF support on SATA disks for the family.  This might be something that was a special for IBM, or an Itanium-only version of their BIOS.  It doesn't make what I posted incorrect, but does change the dynamics to, yes, it exists, but it wont-work-on-your-hardware.

If I had to guess, it is probably likely this is the situation.
PowerEdgeTechIT ConsultantCommented:
Interesting ... is this LSI's 1068E that the PERC is based on, or is this an actual PERC?  Also, is this the non-RAID/JBOD-capable version (equivalent to Dell's SAS 6/iR) or the RAID version (equivalent to Dell's PERC 6/i)?  As far as I can tell, both are based on the 1068E, but only the SAS 6/iR (in non-RAID mode) has been "rumored" to allow >2TB disks (until possibly now).

Author ... sorry if we are getting off-topic ... if you were not considering >2TB drives at all in the first place, then this may not be of any benefit to you.
The PERC6 is a Dell branded version of a family of LSI Controllers that offer a wide range of external/internal port combinations all based on that chipset.  Firmware is effectively the same, differences are more towards branding and disabling some features.

I have hooked up a 3TB SAS to my PERC6 before and it worked fine under solaris, but I did update firmware to a revision that Dell may or may not officially support.   I really just don't remember. But my Dell-branded PERC6 that does support a 3TB SAS is running

It ID's as "RAID SAS1068E B3" MPT 105, MPTBIOS- (2008.08.06)
String is "SAS6IR" which confirms PERC6/iR

Again, when you look at the limitations, there is a hard chip limit on SATA drives that are 512-byte sectored.  That can't be changed because that is locked in the chip.  SATA w/ 4K sectored drives is unverified but that is what Hitachi is claiming on their data sheet. My code does have support for 4K sectored drives but don't have one > 3TB I want to play with right now).

The controller families have supported READ/WRITE(16) and all the other 16-byte CDBs required for targets > 2TBs for years now, so all the hard constraints are limited to the world of SATA.
Dale303Author Commented:
All good info so was happy to let the thread progress. I just got the one drive for now and plan for the size upgrade when there's a bit more free time.
Dale303Author Commented:
All great info. Thanks to all that participated.
Question has a verified solution.

Are you are experiencing a similar issue? Get a personalized answer when you ask a related question.

Have a better answer? Share it in a comment.

All Courses

From novice to tech pro — start learning today.