Drive offline after power down

3.4K Views
Last Post 16 April 2016
Topic Is Solved

nick pomf posted this 13 April 2016

One of the drives in my Promise Pegasus hasn't come back after a power down. This has happened to me before and its infuriating. In the past I've used the command line utility to bring it back online but this time its not working:

cliib> phydrv

===============================================================================

PdId Model Type Capacity Location OpStatus ConfigStatus

===============================================================================

1 Hitachi HDS7 SATA HDD 1TB Encl1 Slot1 OK Array0 No.0

2 Hitachi HDS7 SATA HDD 1TB Encl1 Slot2 OK Array0 No.1

3 Hitachi HDS7 SATA HDD 1TB Encl1 Slot3 OK Array0 No.2

Lost Physical Drives

===============================================================================

PdId Model PhyCapacity Location OpStatus ConfigStatus

===============================================================================

4 ATA 0Byte Unknown Missing Array0 No.3

cliib> phydrv -a online -p 4

Error (0x49841602): invalid physical drive id in data buffer

Any ideas what to do?

14 Comments

Order By: Standard | Latest | Votes

Joe Engledow posted this 13 April 2016 - Last edited 13 April 2016

Please use the phydrv command with extreme caution. We don't recommend using this command without first consulting promise technical support.

Phydrv -a online -p # takes a drive which is offline and has stale data and forces it to be online with the rest of the drives, a risky maneuver that can cause corruption of data and in extreme cases when forcing online a broken drive repeatedly, can cause damage to the Pegasus backplane.

The drive is usually marked offline/dead for a reason and it should almost always be replaced when it happens. Please replace the drive with a new one from the compatibility list, and begin the rebuild from the disk array menu of Promise Utility.

For further assistance, please register your Pegasus under your email address at our eSupport site, http://support.promise.com ; then open a Web Support Case under the registered Pegasus serial number if you have a problem like this. We would be happy to help you over the phone as well Monday-Friday 9-5 Pacific Time.

Joe Engledow, Technical Support, PROMISE Technology (408) 228-1500

nick pomf posted this 13 April 2016

Great. This happens pretty much every time I power this thing down. So you think I should buy a load of spare drives then? One for each time I go on holiday.

There's no chance of stale data on the drive - the whole thing was powered down. How do I add it back please?

Jerry Case posted this 13 April 2016

Hi Nick,

If the drive is coming up missing everytime you turn off the unit and turn it back on, it could be a few things:

1) the drive is defective and not powering up and being detected when Pegasus starts.

2) the drive has poor connection in the slot and comes loose on occasion. Check that it's latched pulled the drive into the slot firmly and not that the drive is unlatched but pushed all the way in. Even unlatched but inserted all the way, it may still be detected until it comes loose again.

3) the slot has a connection issue

4) firmware is out of date on Pegasus. We made improvements in the firmware for drive detection issues for some drive models.

If the drive does not show with the 'phydrv' command then the chassis is not detecting the drive. Unless the drive is detected, no utility built into the Pegasus will bring it back online or repair the array by using the undetected drive.

If the drive had been detected, and as you stated, no change was done to the disk array data by writring to it between the time it was off and back on, then the phydrv -a online command would be safe. if the drive had been dead for some small ammount if time and you had written to the Pegasus storage, then force the disk online, you will potentially corrupt your data and it could possibly be irreversible.

Knowing this, if the unit is powered off and all the disks are unlatched and you try the disk from slot 4 into one of the other slots and then turn it on with the rest of the disks out, if the disk is detected, the array will state there is missing disks. Don't accept missing disks. This also will let you know the disk is at least working for now. I would try the same thing with a known good disk in slot 4 to see if it's detected everytime.

If the drive is undetected in any slot, it is a dead drive.

if you would like troubleshooting assistance in narrowing down the issue, please contact Promise Technical Support directly.

There is a contract fee for technical support only (no hardware replacement) on out of warranty Pegasus systems.

Best Regards

Jerry - Promise Technology

nick pomf posted this 13 April 2016

Sorry, I wasn't clear. This has happened over again, but it's not the same drive each time. Each time it happens I just remount the drive using the "phydrv -a online -p x" command. Except for last time, when it didn't work. I raised a support ticket once and they told me to replace the drive. Which I did, but... The only reason I listened to them was because I was desperate. I really don't believe there was anything wrong with it because this happens so often. In that occasaion, (and the only occasion I've ever opened a support ticket), your support person clearly didn't read what I'd written and they didn't look at the screen shots I sent them. The answer I received was terrible but I had to follow their instructions because I was in a bind.

Do you really think its realistic that I've bought 5 bad drives in 3 years? I hardly use my pegasus, its mostly turned off. I only use it after taking lots of photos. I can't believe that all these drive have failed on me.

I'm 100% sure that there's no risk of corruption becase the drive was turned off. Can you please tell me how to remount drive 4?

Jerry Case posted this 13 April 2016

Unless drive PD 4 is being detected in the chassis, there is no way to force it back online. It is a dead drive.

Try it in another slot to see if it is detected. Try another drive in slot 4 to see if it's detected. if nothing is detecting in slot 4, it could be the slot.

The nature of your issue with the disks coming up dead randomly and often sounds like the firmware on the Pegasus is out of date. What firmware is on the unit?

What drive model are you using? Are these the drives that were provided with the Pegasus or did you upgrade with your own purchased drives?

Best Regards

Jerry - Promise Technology

nick pomf posted this 14 April 2016

Thanks for the reply, I'll try swapping the drives around.

One of the drives is a replacement, like I said, one time this happened your support staff told me to replace it.

How do I determine the version of the firmware?

nick pomf posted this 14 April 2016

Ok, I swapped the drives around. No change, the dead drive is still showing dead, so I've bought another replacement. Worryingly the drive which is now in bay 4 (where the dead one was) now has a flashing blue light where as the other 2 working drives the light is solid. What does this mean? And the Promise Utility is not starting now.

As I'm writing this the light stopped flashing and went out completely. I hope its now dead as well now. That means I've lost my data right?

I tried downloading the latest from here http://www.promise.com/DownloadFile.aspx?DownloadFileUID=4122 but the file (R_SI_504000057.img) won't open on a mac. I get this error:

Jerry Case posted this 15 April 2016

If you have RAID 5 configured for your logical drive, then you can be one disk down and still have the data be readable.

What you probably saw was the replacement disk, when coming from the manufacturer, will often have a partition already on the disk. If this is the case, the disk will show up in the list of available disks as 'pass thru'

Pass through just presents the disk to the computer to be used as a regular disk drive and not being handled by the Pegasus

You have to change any Pass Thru drives to unconfigured before the disk will be used to rebuild an array.

Even with one disk down, three disks online and the array online the disk array should list as degraded and the logical drive as critical. Critical because is you lose one more disk, then you do lose your data by any regular means of reading it. However a data recovery service may be able to get the data back for a price.

As with any hard disk drive, disk array or any storage for that matter, it is a good idea to have back-up of the data.

RAID levels 1 and higher have some level of fault tolerance but that's just tolerance, not fail proof. It won't protect from drive fails greater than the design of the RAID can lose and it is not protected for any kind of disaster.

You said the utility won't start now. It won't start if it's not able to detect the Pegasus. So with it attached, go to the Mac's terminal screen and in the window, type 'promiseutil' like you had before to get to the first phydrv command.

if you start the promise utility from the terminal window, type 'phydrv' to list all the disks. Drives 1, 2 and 3 should be unchanged as your first message. Drive 4 should be detected but then you can see if it's on pass thru or not.

Also, the firmware update is not a disk image to mount on the Mac. It's an image file passed to the Pegasus through the utility to install.

However our latest update is now available on automatic updates, so just launch your utility and check for updates.

Best Regards

Jerry - Promise Technology

nick pomf posted this 16 April 2016 - Last edited 16 April 2016

Thanks for the reply, but I'm not sure you understand. I swapped the disks in bays 3 & 4 as you suggested to see if there was a problem with one of the bays. Now the 2nd blue light isn't coming on for either drive. I've put them back again, and the phydrv command is showing this still:

cliib> phydrv

===============================================================================

PdId Model Type Capacity Location OpStatus ConfigStatus

===============================================================================

1 Hitachi HDS7 SATA HDD 1TB Encl1 Slot1 OK Array0 No.0

2 Hitachi HDS7 SATA HDD 1TB Encl1 Slot2 OK Array0 No.1

3 Hitachi HDS7 SATA HDD 1TB Encl1 Slot3 OK Array0 No.2

Lost Physical Drives

===============================================================================

PdId Model PhyCapacity Location OpStatus ConfigStatus

===============================================================================

4 ATA 0Byte Unknown Missing Array0 No.3

I've now repaced the drive in bay 4 with a new one. Is there a good example online that shows how to rebuild the array? I can't click 'rebuild' as the button isn't enabled:

Joe Engledow posted this 16 April 2016 - Last edited 16 April 2016

The array status is still incomplete if there are lost drives showing up.

Accept the array's missing drives with the Accept button in Disk Array. The array status should be Degraded after that.

Insert the new drive hot and wait 1 minute.

If it is a drive from the store it will have its own 1-drive partition so the Pegasus will put it up as a passthru. In the settings of the physical drives individually, change the status of the new drive #4 from "PassThru" to "Unconfigured" and save it. It will erase the drive's existing partition so you will be prompted to confirm this erasure. Once the new physical drive is "Unconfigured", then go to Disk Array and begin the Rebuild.

nick pomf posted this 16 April 2016

I'm a bit nervous about this. Firstly please confirm that I should press accept in the disk array menu? The warning message says this can result in lost data.

Secondly, as I stated the new drive is already in bay 4, so I cannot "Insert the new drive hot and wait 1 minute." because its already in there. I assume that's ok?

What do you mean "If it is a drive from the store"? What store?

Joe Engledow posted this 16 April 2016 - Last edited 16 April 2016

The warning is meant to sound dire because if more than one drive is missing then it will go badly, it's intended so people will check with Promise Support. We've assured you through this forum thread that the array and logical drive will be online with 3 physical drives. If you do not perform this step, we cannot proceed.

If you've already inserted the drive then please leave it in.

I wrote "from the store" to differentiate it from "drives you buy from Promise" which are already unconfigured.

Joe

nick pomf posted this 16 April 2016

Ok, thanks. That's rebuilding now. Fingers crossed.

nick pomf posted this 16 April 2016

Thanks for your help.

Drive offline after power down

Search

Categories