2 x raid 1 with four disks

1. offizieller Beitrag

    Dear Community,

    I have a HP ProLiant microserver Gen8 and used two disks so far (4tb each). I decided to upgrade to four disks and now I have options with respect of RAID.


    Can I add another 2 disks in Raid 1 with a bigger HDD capacity (i.e 8 Tb each)? I would like to avoid to create a RIAD10 in order to have more space, always if I can have a RAID1 of 4tb (4tb + 4 tb) and a RAID of 8tb (8tb + 8 tb) in my ProLiant microserver Gen8 at the same time. Is there any disadvantage to have two RAID1 with different capacity?


    Thank you very much!

    Thank you very much for the confirmation.


    Actually, reading here and there, I understood that most probably it could be better for me going with to no raid and no pool and use rsync to create a backup. Or another option could be use UnionFS and Snapraid or no pool and MergeFS. I have to understand better as I am just learning and I must study more. Please could you explain me:

    1) Can I use snapraid to create a backup/copy instead than rsync. I mean, Is snapraid a substitute of rsync to backup data? Any disadvantage or advantage?

    2) Is snapraid or no pool fast as RAID0 or RAID1? Which one is better in my case?

    3) If I go for no pool or snapraid, should I choose EXT4 or BTRFF? Any disadvantage or advantage?

    4) If with UnionFS and snapraid I can create a copy/backup of the files, what option I have to copy/backup files in margeFS? Is Rsync an option?

    5) With MergeFS, UnionFS and snapraid, can I use HDDs with different capacity?


    I am sorry for all these questions and if I am doing some confusion, I just took info here and there and I am trying to connect in order to make the best choice for my server. If you have any advice on how I should proceed, I am ready to learn.


    Thank you!!

  • macom

    Hat das Label gelöst hinzugefügt.
    • Offizieller Beitrag

    Before you even consider Snapraid I would suggest looking at the website there is a lot of information on there, in respect of mergerfs


    The plugins install each and allow you to configure them from OMV's GUI, but that is as far as it goes, you would get some support in here, mainly from users already using both. So you have to understand how both work.


    Question 3, use ext4, Question 5, Yes


    As far as Raid is concerned we usually ask why do you need to use it, with raid1 the data is written to both drives, so both are spinning to facilitate that. By using two drives one for data and one as a backup you get, in essence, the same as raid1, the only difference being is that the backup drive can be spun down when not in use.


    There is also a New User Guide contains a lot of useful information

    Thank you so much to have found the time to answer to my questions and share links where I can find useful info, a lot appreciated -||-. I will try to understand better, as my ideas are still not clear enough.


    Honestly I decided to go with RAID because I did not have good knowledge (and still I do not have ahahah) and I thought it was the right thing to do with a server in order to access to the data quickly and with redundant data. I thought it was the only and best file system for a server. As I have to expand the capacity of my server now, I did some research and here I am with my doubts :). By the way something is sure, I will replace the RAID with a "normal" one (I do not know the technical nomenclature), if I need some pool in the future I could use the snapraid (still I have to read from your link).


    Before I start to read the links above, to better understand the direction to focus to, please could I ask further if the reason to mergerfs? as I cannot understand the utility to marge two or more folders, maybe because my folders are not so big in dimension.


    If I did not understood wrong, you are advice to start making two drives one for data and one for backup. Should I use rsync to backup the data I need to save?


    Cheers -||-

    • Offizieller Beitrag

    mergerfs is used to merge two or more filesystem so that they look as if they were just one big filesystem



    Zitat von FROMV

    If I did not understood wrong, you are advice to start making two drives one for data and one for backup. Should I use rsync to backup the data I need to save?

    rsync is one possibility which gives you a 1:1 copy. Another option is rsnapshot which is using rsync in the background. Advantage is that you can get versioned backups without using much space on the disk as only new files take up additional space

    So with rsnapshot you really get a bonus over raid 1. If you delete a file and run rsync the file is gone on both drives (depending on whether or not you are using the delete option). With rsnapshot you still have the copies of this file in your previous snapshots.

    You can also go back in time and restore previous versions of a file.

    Thank you.

    Indeed, rsnapshot is a very useful tool. Most probably I could save my job, as I have replaced with a corrupted file, but of course it was too late after 10 sec I copied the file :(.


    1) If I go for normal file system (1 disk for data and 1 disk for backup), do I have SMART monitor, power control and check/checksum tool available? (I have read all the RAQ of snapraid)?

    2) mergerfs is also for disk? Of course I can share the maim folder (an do all the files and folders inside), but does it have some advantage over UnionFS if I have to marge two or more disks?

    3) The new HDDs are arriving :), which one is the best method to copy all the files from RAID1 (4 TB) to the new disks? Is there any good tip to follow in order to check the integrity of the files? Essentially I mean if there is a way to stress less the HDD during the copy and keep or check the integrity of the files?


    Thank you again for the help!!

    • Offizieller Beitrag

    1) SMART, power control (APM) yes, checksum no; for this you would need an additional tool

    2)the UnionFS plugin uses mergerfs in the background

    3) rsync; if you do a second run with the --checksum flag you can test if everything was copied correctly (the run with the --checksum flag should not do anything)

    https://linux.die.net/man/1/rsync

    Thank you so much for your sharing knowledges, I know, it can sound repetitive, but I really appreciate your help guys. I have learnt something today as well and it is a day well spent :)!!!


    1) The additional tool for checksum you are talking in point 1) is rysnc as for point 3), isn't it? Or were you advising anther solution which you prefer?

    2) For SMART monitoring I have just to enable and use the the one in OMV GUI, isn't it? No further action required?

    3) Reading the checksum in the rsync link you kindly shared, I am not sure I understood well, as it seems that the checksum is enabled before the first transfer. By the way if I understood well your advice, I should make the first transfer without the checksum enabled (rsync is doing a kind of checksum by default). Just when the transfer is fully finish I have to make a second scan with the checksum enabled. Is it correct?

    4) After the first copy should I keep the checksum enabled or do you suggest to switch it off after the the second run?

    5) From the instructions "Generating the checksums means that both sides will expend a lot of disk I/O reading all the data in the files in the transfer (and this is prior to any reading that will be done to transfer changed files), so this can slow things down significantly." Does it mean that the space will be reduced generating the checksum and the speed of my HDDs slowed down?

    6) In your experience, do you have any advice for the checksum? Always open to learn from you guys!


    Thanks!!!

    • Offizieller Beitrag

    1) I use btrfs as filesystem

    2) yes

    3)

    Code
    -c, --checksum              skip based on checksum, not mod-time & size

    Without --checksum, rsync will use size and modification time to identify changed files that need to be copied.

    With --checksum, rsync will not use modification time and size, but the checksum to identify changed files.

    4) no. The suggestion would be to run it only once for the first copy if you worry that something might have gone wrong and some of the data might be corrupted

    5) should be only an issue for the verification run, as you can switch it off after that

    6) see above, maybe somebody else has a different opinion

    Is snapraid to prefer to normal data disk and backup, considering checksum and scrubbing? Is it giving any advantage over the normal data disk and backup? It is something that I cannot understand.


    Furthermore how parity space works? If I have 3 data disks of 6+4+4tb (14tb), is 6 tb of parity enough? The parity is like a backup or it take less space than normal backup with equal amount of data?


    Hopefully my questions are not too silly.


    Thanks!!

Jetzt mitmachen!

Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!

Benutzerkonto erstellen Anmelden