List Readers -
I have a Dell server that uses the Perc 6i controller and had 5 1Tb disks installed (1 for OS and the other 4 in a Raid0 for a large storage pool). The owner of the server wanted me to swap out the 1Tb disks for 2Tb disks - easy enough I thought, but I ran into some issues trying to clone the OS disk to the new 2Tb disk, so I just did a re-install. So basically we now have 5 2Tb disks (1 for OS and the other 4 in a Raid0) like before. I was able to copy off the data that was on the original Raid0 volume and put it back on the newly created one that is now 8Tb and that worked well enough. But since I was unable to clone the original OS disk, and proceeded to simply re-install the OS, the owner apparently wants his $HOME directory copied off (which of course resides on the 1Tb OS disk). So, being a SAS disk, I purchased an ICY Dock which allows me to mount the SAS disk in an external enclosure, and plug it in via USB (SAS -> SATA -> USB). However when I do a 'fdisk -l /dev/sdc' I receive the following:
[root@gc-server2 ~]# fdisk -l /dev/sdc
Disk /dev/sdc: 2199.0 GB, 2199023255552 bytes
255 heads, 63 sectors/track, 267349 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk /dev/sdc doesn't contain a valid partition table
Two things are a bit unusual here, first of all its not a 2Tb disk, and secondly why cant I read the partition table.
Originally the OS was on an LVM volume and by default it was setup as
/dev/VolGroup00/LogVol00 / ext3 defaults 1 1
My question is, is there a way to connect and mount this via USB and be able to get the $HOME directory stuff off that I failed to copy off before removing the disk?
Thanks in advance
Michael
===================================
P Please consider the environment before printing this e-mail
Cleveland Clinic is ranked one of the top hospitals in America by U.S.News & World Report (2009). Visit us online at http://www.clevelandclinic.org for a complete listing of our services, staff and locations.
Confidentiality Note: This message is intended for use only by the individual or entity to which it is addressed and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If the reader of this message is not the intended recipient or the employee or agent responsible for delivering the message to the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please contact the sender immediately and destroy the material in its entirety, whether electronic or hard copy. Thank you.
Weiner, Michael wrote, On 05/24/2010 10:28 AM:
List Readers -
I have a Dell server that uses the Perc 6i controller and had 5 1Tb disks installed (1 for OS and the other 4 in a Raid0 for a large storage pool).
[root@gc-server2 ~]# fdisk -l /dev/sdc
Disk /dev/sdc: 2199.0 GB, 2199023255552 bytes
255 heads, 63 sectors/track, 267349 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk /dev/sdc doesn't contain a valid partition table
Are you sure that the Perc was not presenting you with a logical disk (Volume)? One I received from dell, with RHEL on it, was setup with 8 disks in a raid5 and the controller was presenting the whole array as two drives, i.e., no physical drives are presented to the OS.
BTW had your support contract ran out with Dell/RH?
Two things are a bit unusual here, first of all its not a 2Tb disk, and secondly why cant I read the partition table.
Originally the OS was on an LVM volume and by default it was setup as
/dev/VolGroup00/LogVol00 / ext3 defaults 1 1
you might get lucky with a vgscan or lvmdiskscan, however my bet is on having to put all the drives back in a perc array.
My question is, is there a way to connect and mount this via USB and be able to get the $HOME directory stuff off that I failed to copy off before removing the disk?
Thanks in advance
Michael
Todd Denniston, On Monday, May 24, 2010 11:15 AM
Are you sure that the Perc was not presenting you with a logical disk
(Volume)?
One I received from dell, with RHEL on it, was setup with 8 disks in a
raid5 and the controller was
presenting the whole array as two drives, i.e., no physical drives are
presented to the OS.
This is a possibility. I will have to check out the other server hanging around
BTW had your support contract ran out with Dell/RH?
We do have support on the box, but don't purchase a RHEL license and we run CentOS instead
you might get lucky with a vgscan or lvmdiskscan, however my bet is on
having to put all the drives
back in a perc array.
I will try the lvmdiskscan, but vgscan only sees the currently running disk. Problem is, if I put it back into the Perc then it will want to 'initialize' it which I believe will do a low level format :(
Thanks Michael
===================================
P Please consider the environment before printing this e-mail
Cleveland Clinic is ranked one of the top hospitals in America by U.S.News & World Report (2009). Visit us online at http://www.clevelandclinic.org for a complete listing of our services, staff and locations.
Confidentiality Note: This message is intended for use only by the individual or entity to which it is addressed and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If the reader of this message is not the intended recipient or the employee or agent responsible for delivering the message to the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please contact the sender immediately and destroy the material in its entirety, whether electronic or hard copy. Thank you.
Weiner, Michael wrote, On 05/24/2010 11:38 AM:
Todd Denniston, On Monday, May 24, 2010 11:15 AM
BTW had your support contract ran out with Dell/RH?
We do have support on the box, but don't purchase a RHEL license and we run CentOS instead
you might get lucky with a vgscan or lvmdiskscan, however my bet is on
having to put all the drives
back in a perc array.
I will try the lvmdiskscan, but vgscan only sees the currently running disk. Problem is, if I put it back into the Perc then it will want to 'initialize' it which I believe will do a low level format :(
With the Dell support, you could call Dell and ask how they expect the Perc to work with the disk upon reinsertion. Should at least letup your nerves.
Todd Denniston, On Monday, May 24, 2010 11:50 AM To: CentOS mailing list
With the Dell support, you could call Dell and ask how they expect the
Perc to work with the disk
upon reinsertion. Should at least letup your nerves.
Thanks Todd, I think I will have to resort to that :)
Michael
===================================
P Please consider the environment before printing this e-mail
Cleveland Clinic is ranked one of the top hospitals in America by U.S.News & World Report (2009). Visit us online at http://www.clevelandclinic.org for a complete listing of our services, staff and locations.
Confidentiality Note: This message is intended for use only by the individual or entity to which it is addressed and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If the reader of this message is not the intended recipient or the employee or agent responsible for delivering the message to the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please contact the sender immediately and destroy the material in its entirety, whether electronic or hard copy. Thank you.
On Mon, May 24, 2010 at 10:28 AM, Weiner, Michael weinerm@ccf.org wrote:
List Readers –
I have a Dell server that uses the Perc 6i controller and had 5 1Tb disks installed (1 for OS and the other 4 in a Raid0 for a large storage pool). The owner of the server wanted me to swap out the 1Tb disks for 2Tb disks – easy enough I thought, but I ran into some issues trying to clone the OS disk to the new 2Tb disk, so I just did a re-install. So basically we now have 5 2Tb disks (1 for OS and the other 4 in a Raid0) like before. I was able to copy off the data that was on the original Raid0 volume and put it back on the newly created one that is now 8Tb and that worked well enough. But since I was unable to clone the original OS disk, and proceeded to simply re-install the OS, the owner apparently wants his $HOME directory copied off (which of course resides on the 1Tb OS disk). So, being a SAS disk, I purchased an ICY Dock which allows me to mount the SAS disk in an external enclosure, and plug it in via USB (SAS -> SATA -> USB). However when I do a ‘fdisk –l /dev/sdc’ I receive the following:
[root@gc-server2 ~]# fdisk -l /dev/sdc
Disk /dev/sdc: 2199.0 GB, 2199023255552 bytes 255 heads, 63 sectors/track, 267349 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes
Disk /dev/sdc doesn't contain a valid partition table
Two things are a bit unusual here, first of all its not a 2Tb disk, and secondly why cant I read the partition table.
Originally the OS was on an LVM volume and by default it was setup as
/dev/VolGroup00/LogVol00 / ext3 defaults 1 1
My question is, is there a way to connect and mount this via USB and be able to get the $HOME directory stuff off that I failed to copy off before removing the disk?
Thanks in advance Michael
Since the disk was connected through a RAID controller it's most likely that the disk is formatted for the custom format that the RAID controller uses. The controller then presents that disk to the OS. You're going to have to put the drive back into the server so the PERC can see it, then you might be able to get to the LVMs.
PS. You did something like this without making backups first?
PPS. You might have a reason, but it would be irresponsible not to point out that by using RAID0 you now have an exponential risk of losing all of that data.
Brian Mathis, On Monday, May 24, 2010 11:17 AM
Since the disk was connected through a RAID controller it's most likely that the disk is formatted for the custom format that the RAID controller uses. The controller then presents that disk to the OS. You're going to have to put the drive back into the server so the PERC can see it, then you might be able to get to the LVMs.
You are probably right, but I believe it will want to initialize the drive first which I cannot let it do as I believe that will wipe whats there.
PS. You did something like this without making backups first?
The only thing I really cared about was the data on the Raid0 honestly and I wasn't too concerned about the $HOME directory since there is another 'copy' on another server - however the individual has since let me know that they were not in sync and there were things he had done that he hadn't backed up.
PPS. You might have a reason, but it would be irresponsible not to point out that by using RAID0 you now have an exponential risk of losing all of that data.
The Raid0 is simply a large disk pool for temporary data storage. Not the way I would build this, but I am simply managing the hardware for this individual, and he has been warned about the loss of data due to the single point of failure, especially without backing this data up to somewhere I can get to it with tape
Thanks again Michael
===================================
P Please consider the environment before printing this e-mail
Cleveland Clinic is ranked one of the top hospitals in America by U.S.News & World Report (2009). Visit us online at http://www.clevelandclinic.org for a complete listing of our services, staff and locations.
Confidentiality Note: This message is intended for use only by the individual or entity to which it is addressed and may contain information that is privileged, confidential, and exempt from disclosure under applicable law. If the reader of this message is not the intended recipient or the employee or agent responsible for delivering the message to the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited. If you have received this communication in error, please contact the sender immediately and destroy the material in its entirety, whether electronic or hard copy. Thank you.
On Mon, May 24, 2010 at 11:43 AM, Weiner, Michael weinerm@ccf.org wrote:
Brian Mathis, On Monday, May 24, 2010 11:17 AM
...
PPS. You might have a reason, but it would be irresponsible not to point out that by using RAID0 you now have an exponential risk of losing all of that data.
The Raid0 is simply a large disk pool for temporary data storage. Not the way I would build this, but I am simply managing the hardware for this individual, and he has been warned about the loss of data due to the single point of failure, especially without backing this data up to somewhere I can get to it with tape
Thanks again Michael
...
OK, I won't beat this to death, as I know there are always exceptions to the rules, BUT a 4-way RAID0 is far more likely to fail than just a single disk, so I wouldn't consider it a single point of failure. I'm not sure about the math right now, but it's something like 2^4 (16) or even 4! (24) time more likely to fail than a single disk.
-
Brian Mathis -
Rob Kampen -
Todd Denniston -
Weiner, Michael