Jump to content

Just Me

Members
  • Posts

    107
  • Joined

  • Last visited

Posts posted by Just Me

  1. Hey!

     

    Sorry to bother you guys but I'm a little bit insecure when I reconfigure my server and I would like a sanity check if everything would work the way I've planed it.

     

    My current setup:

    unraid 6.7.2

    1x 10TB single parity

    10x data drives (1-3 TB size)

     

    What I would like to do:

    Add another 10 TB drive, encrypt this drive, remove some smaller drives and reorder my disks within my array (of course I want to keep my data 😃)

    Though I'm not sure if encryption is the best choice since my antique CPU doesn't have AES instructions. So the array read / write speed may be slower.

    Well, we will see ^^

     

    The new drive is precleared and SMART data look fine.

     

    0. Backup my flash

    1. Stop the array

    2. Assign the new 10 TB drive as data disk 11

    3. Change the file system for disk 11 to xfs encrypted

    4. Enter a passphrase and start the array

    5. Format the drive

     

    Now I should have an empty, encrypted data drive. I want to get rid of disk 2, 3 and 9 since they are the smallest drives.

     

    6.1 Copy all data from disk 2 to disk 11 (using screen): rsync -avPX /mnt/disk2/ /mnt/disk11/

    6.2 Copy all data from disk 3 to disk 11 (using screen): rsync -avPX /mnt/disk3/ /mnt/disk11/

    6.3 Copy all data from disk 9 to disk 11 (using screen): rsync -avPX /mnt/disk9/ /mnt/disk11/

     

    All data should be on disk11 now, so I could remove the no longer needed drives from the array.

     

    7. Stop the array

    8. Use tools -> new config to reset the array

     

    I'm not really sure what option I should choose regarding "Preserve current assignments". I assume since I want to reorder my drive and remove some disks I could select parity slots only.

     

    9. Select "Preserve current assignments: parity slots" and check "Yes I want to do this", then hit apply.

    10. Back to main, parity should still be selected, all other disks marked as "unassigned".

    11. Check if parity is really really really the correct drive 😉

    12. Assign the data disks in the order I prefer, drives I want to remove will stay unassigned. (The drives can be assigned in any order, or are there any restrictions?)  

    13. Start the array.

     

    Since parity is no longer valid, parity will be rebuild. Approximately 26 hours later everything should be fine and the array will be "protected" again.

     

    Am I missing anything? Any mistakes?

     

    Thank you in advance.

  2. 1 hour ago, johnnie.black said:

    Dual parity should be OK, depending on the number of data disks

    Thank you. The server is designed for 12 disks, so there would be 10 data disks at max. You think that would be OK?

     

    1 hour ago, johnnie.black said:

    since that CPU doesn't support AES instructions.

    Is there an AM3 socket CPU that supports AES instructions?

     

    13 minutes ago, drdebian said:

    It certainly won't leave room for other CPU-intensive tasks, but for file storage you should be fine.

    Thank you. 80+ MB/s seems OK. The server is just used for media and backups, no docker or VMs, mostly vanilla unRAID.

    Maybe I will give it a try, what could possibly go wrong 😂

     

    PS: Stupid question: Does ten encrypted data drives need more CPU performance as just one or two?

     

  3. Hey!

     

    I'm running unRAID 6.7.2 on pretty old hardware:

     

    CPU: AMD Athlon II X2 235e @ 2700 MHz

    Motherboard: ASUS M4A785T-M

    RAM: 12 GB DDR3-1333 ECC

    HBA: Dell H310 & onboard SATA

     

    I would like to switch the file system from xfs to xfs encrypted. But I'm not sure if this old CPU can handle the encryption. Will the use of an encrypted file system slow down the read and write speed of the array in a noticeable way? Or is the CPU fast enough to handle the encryption?

     

    Same question for dual parity, as far as I know a dual parity configuration is much more demanding as single parity.

     

    Thank you in advance for a reply.

  4. 3 hours ago, itimpi said:

    A pending sector can also disappear if the next attempt to write to it succeeds.    It only gets reallocated if that also fails.

    Thank you, new to me. Good to know.

     

    3 hours ago, johnnie.black said:

    That's the best option, SMART attributes some times behave strangely, possible due to firmware issues, extended test will confirm if there's still a problem or not. 

    The extended test finished without errors. SMART data look fine. Still two read error in the syslog though.

     

    Maybe just reboot and check if there are still errors? Or try a non correcting Parity-Check? Not really sure how to handle these errors 🤷‍♂️

  5. Hey!

     

    Yesterday unRAID 6.7.2 send me a warning that one drive has current pending sectors, I've checked the SMART data at disk 3 and the drive had one current pending sector. Since it was late, I just shut down the server to deal with the problem today. After I've started the server the current pending sector has gone, but the reallocated sector count is still zero, offline uncorrectable is also zero. How can a current pending sector just disappear without being corrected as reallocated sector?

     

    The syslog shows two read error on disk 3 though but the SMART data are fine (again).

    Jul  5 11:30:45 NAS kernel: print_req_error: critical medium error, dev sdi, sector 1465157503
    Jul  5 11:30:45 NAS kernel: md: disk3 read error, sector=1465157440
    Jul  5 11:30:45 NAS kernel: md: disk3 read error, sector=1465157448

    I'm running a SMART extended self-test right now, maybe the test will show a problem.

     

    But I'm still confused about this behavior, I don't get it how a pending sector can just disappear 🤷‍♂️

     

    Any advice how I should continue if the extended SMART test shows no problems?

     

    I've attached the diagnostics and the current disk 3 SMART report.

    nas-diagnostics-20190705-1140.zip WDC_WD15EARS-00Z5B1_WD-WMAVU1156825-20190705-1344.txt

  6. Thank you very much. I'll give it a try as soon as my external drive arrives. Preclear is always a good way to "burn in" new drives, even if I don't plan to use it with unRAID.

     

    SATA or eSATA would be better but the device only has USB 3.0 and I don't want to open the external case, nowadays this always results in a destroyed case because of all these plastic clips :-/

  7. Sorry, if this has been asked before, but is it also possible to preclear an external drive attached via USB or is preclear only able to use disks attached via SATA? 

  8. Hey!

     

    I'm using unRAID 6.3.5 and I would like to use my yahoo account to send notifications. But after I've set up my account in unRAID the test fails: Test result: Authorization failed.

     

    The system log says that username or password are invalid:

    Sep 25 15:29:25 NAS sSMTP[13390]: Creating SSL connection to host
    Sep 25 15:29:25 NAS sSMTP[13390]: SSL connection using ECDHE-RSA-AES128-GCM-SHA256
    Sep 25 15:29:28 NAS sSMTP[13390]: Authorization failed (535 5.7.0 (#MBR1212) Incorrect username or password.)

     

    But I'm 100% sure, that user and password are correct. I've tried copy & paste and I've tried to type it in three times. 

    Are there any character that are prohibited in the password?

     

    The help says that the password is stored unencrypted in the email configuration file. Where do I find this file to check the stored password?

     

    Thank you in advance. 

  9. So I've tried to use the command line to preclear my USB disk but it seems it is not that easy. If I start the preclear script the following error appears:

    Pre-Clear unRAID Disk /dev/sda

    ################################################################## 1.15c

     

    smartctl may not be able to run on /dev/sda with the -d ata option.

    however this should not affect the clearing of a disk.

    smartctl exit status = 2

    smartctl 6.2 2013-07-26 r3841 [x86_64-linux-4.1.18-unRAID] (local build)

    Copyright © 2002-13, Bruce Allen, Christian Franke, www.smartmontools.org

     

    Read Device Identity failed: Invalid argument

     

    A mandatory SMART command failed: exiting. To continue, add one or more '-T permissive' options.$

    Do you wish to continue?

    (Answer Yes to continue. Capital 'Y', lower case 'es'):

     

    Of course I could prelcear the disk anyway but the script won't be able to grab the SMART data. I think the issue is the "-d ata" option. The USB device needs a "-d sat" command to read the SMART attributes.

    Maybe I could modify the preclear script to use the "-d sat" option instead of "-d ata" but the script is very complex and I don't think I'm able to modify it correctly  :-\

     

    Edit: okay, I confess I'm just stupid and didn't recognize that the script already allows to set a custom device type for SMART. My bad  :-X:-\

  10. Thanks for the quick reply. It seems the plugin is not able to use USB drives, the attached device is not listet as a valid disk to preclear the plugin just reports: No unassigned disks available.

     

    My USB drive seems to report SMART attributes, CrystalDiskInfo is able to deliver the same data as shown on internal disks. So it seems SMART report is not in issue.

     

    I'll check if I'm able to start a "manuell" preclear using screen and command line.

     

  11. Is there a way to preclear USB attached drives? I like to "burn in" new disks with preclear even if I don't plan to use them in unRAID. I've got a new external 2 TB 2,5" USB disk and I would like to see if it will survive the first  X hours preclearing before I'll put real data on it.

×
×
  • Create New...