Re: mdadm raid1 lost (my fault)

2018-06-06 Thread Wolfgang Pfeiffer
On Wed, Jun 06, 2018 at 04:30:42PM +0200, Wolfgang Pfeiffer wrote: On Sat, Jun 02, 2018 at 08:51:02PM -0500, page p wrote: tldr; what might save you is if you have a backup of your luks header. See further down here and the cryptsetup FAQ. Plus: I'd ask on the luks list for help. Sorry for a l

Re: mdadm raid1 lost (my fault)

2018-06-06 Thread Wolfgang Pfeiffer
On Sat, Jun 02, 2018 at 08:51:02PM -0500, page p wrote: tldr; what might save you is if you have a backup of your luks header. See further down here and the cryptsetup FAQ. Plus: I'd ask on the luks list for help. Hi, I have two drives in a raid 1 configuration and used luks crypt on the raid

Re: mdadm raid1 lost (my fault)

2018-06-03 Thread page p
On 06/03/2018 11:09 PM, Samuel Sieb wrote: Did you stop the RAID array?  "mdadm -S /dev/md127" What is the output of "mount"? yeah forgot that step... but stopping the array and mounting returns: mount: /dev/sdb1: can't read superblock Given your comments below and since cryptsetup can't seem

Re: mdadm raid1 lost (my fault)

2018-06-03 Thread Samuel Sieb
On 06/03/2018 10:35 PM, page p wrote: On 06/03/2018 11:09 PM, Samuel Sieb wrote: Try "mount -t xfs -o ro /dev/sdb1 /mnt/localhost-sdb1". returns: mount: /dev/sdb1 is already mounted or /mnt/localhost-sdb1 busy Did you stop the RAID array? "mdadm -S /dev/md127" What is the output of "mount

Re: mdadm raid1 lost (my fault)

2018-06-03 Thread page p
On 06/03/2018 11:09 PM, Samuel Sieb wrote: Try "mount -t xfs -o ro /dev/sdb1 /mnt/localhost-sdb1". returns: mount: /dev/sdb1 is already mounted or /mnt/localhost-sdb1 busy perhaps since i --created the array would this format the soft raid but leave the drive data intact not needing to mkfs

Re: mdadm raid1 lost (my fault)

2018-06-03 Thread Samuel Sieb
On 06/03/2018 08:46 PM, page p wrote: [fluorine 10:32:38 ~] $ file -s /dev/sdb1 /dev/sdb1: SGI XFS filesystem data (blksz 4096, inosz 512, v2 dirs) [fluorine 10:32:42 ~] $ mount /dev/sdb1 /mnt/localhost-sdb1 mount: unknown filesystem type 'linux_raid_member' Try "mount -t xfs -o ro /dev/sdb1 /

Re: mdadm raid1 lost (my fault)

2018-06-03 Thread page p
On 06/03/2018 06:39 PM, Samuel Sieb wrote: On 06/02/2018 11:39 PM, page p wrote: file -s just returns : data i since remove the mdadm.conf so it has auto assigned back to 127... Can you show us the exact commands you are trying and the result? https://paste.fedoraproject.org/paste/e5MBmE67

Re: mdadm raid1 lost (my fault)

2018-06-03 Thread Samuel Sieb
On 06/02/2018 11:39 PM, page p wrote: file -s just returns : data i since remove the mdadm.conf so it has auto assigned back to 127... Can you show us the exact commands you are trying and the result? https://paste.fedoraproject.org/paste/e5MBmE67dpPNlgNAX7U7Rw What does "file -s /dev/sdb

Re: mdadm raid1 lost (my fault)

2018-06-03 Thread page p
On 06/03/2018 03:46 AM, François Patte wrote: Le 03/06/2018 à 08:39, page p a écrit : Hi and thank you, file -s just returns : data i since remove the mdadm.conf so it has auto assigned back to 127... Is there any inconvenience to have an array with number 127? not really, just numbering..

Re: mdadm raid1 lost (my fault)

2018-06-03 Thread François Patte
Le 03/06/2018 à 08:39, page p a écrit : > Hi and thank you, > > file -s just returns : data > > i since remove the mdadm.conf so it has auto assigned back to 127... Is there any inconvenience to have an array with number 127? -- François Patte UFR de mathématiques et informatique Laboratoire

Re: mdadm raid1 lost (my fault)

2018-06-02 Thread page p
Hi and thank you, file -s just returns : data i since remove the mdadm.conf so it has auto assigned back to 127... Can you show us the exact commands you are trying and the result? https://paste.fedoraproject.org/paste/e5MBmE67dpPNlgNAX7U7Rw thank you, page On 06/02/2018 11:48 PM, Samue

Re: mdadm raid1 lost (my fault)

2018-06-02 Thread Samuel Sieb
On 06/02/2018 06:51 PM, page p wrote: the two drives can be seen, are part of an array and a hexdump says that they are still XFS but i can not get luks to open the array. If it is supposed to be a LUKS device, then you should not be able to see the XFS filesystem. Can you show us the exact c

mdadm raid1 lost (my fault)

2018-06-02 Thread page p
Hi, I have two drives in a raid 1 configuration and used luks crypt on the raid once created. today i wanted to change the device number from 127 to 0 and can no longer open the array as luks says it is not a valid device. what i did was: $ mdadm --stop /dev/md127 $ mdadm --create --verbos