Jump to content

Umfriend

Members
  • Posts

    1001
  • Joined

  • Last visited

  • Days Won

    54

Posts posted by Umfriend

  1. How did you get it in?

    Some here use DP together with 3rd party parity software, I believe SnapRaid is most popular. Perhaps that would work for you but I would not be able to advise on how to do that.

  2. That ain't easy.

    Best thing for you, IMHO, is to have two Pools, one for the old drives and one for the new. Then copy from old Pool to new. Power off machine and disconnect all old drives.

    I realise this is different from what I suggested before but at that time I was not aware you wanted to have the old drives as sort of a backup. Also, this way the usefullness of the old drives as a backup will fade over time, I would expect, as it does not track changes/new files.

  3. With future products I would say the expectation is they will be there at some point.

    It is not that hard to find non-Helium drives as all helium drives I've seen explicitly advertise this. So if it is not stated then it is not Helium.

    Also, typically, if you can afford it, it is not a Helium-filled drive.

  4. If you use x2 duplication and want to use the SSD Optimizer plug-in, you'll need two SSDs (or play around with hierarchical Pools, then you can do it with one SSD (or at least that has been written here, I actually consider that a bug and maybe it is "solved" but I could tell you how to do it.)). And yes, I imagine that if you do it will make a difference when you upload that much provided that there is no other bottleneck (1Gb Lan for instance).

    On Clouddrive as a backup: No idea.

  5. So I think it is a matter of use case and personal taste. IMHO, just use one Pool. If you're going to replace the 5900rpm drives anyway over time anyway. I assume you run things over a network. As most are still running over 1Gbit networks (or slower), even the 5900rpm drives won't slow you down.

    I've never used an SSD Optimizer plugin but yeah, it is helpful for writing, not reading (except for the off-chance that the file you read is still on the SSD). But even then it would need a datapath that is faster than 1Gbit all the way.

    What you could do is test a little by writing to the disks directly outside the Pool in a scenario that resembles your usecase. If you experience no difference, just use one Pool, makes management a lot easier. If anything, I would more wonder about duplication (do you use that?) and real backups.

  6. 5 minutes ago, gd2246 said:

    Not true, you nailed it! :D

    A bit by accident as I did not really expect that simple thing to make such a difference. Anyway, you are welcome.

    What you say about the "Other", yeah, that is part of it. Donnow whether it can lead to more serious consequences but stopping the DP Service when you want to do your own file tranfers is best practice.

    Now to nitpick, SnapRAID is not a backup. Sure, it may save you from HDD failure to some extent but not against accidental deletions, theft, fire, simultaneous multiple HDD failure (and you got a few). If you value your data, you might consider a real solution. You wouldn't want to lose your personal copy of the entire interweb (91TB, uhm, wow!) :D

  7. BTW, if you move files among pooled-drives, I think it is strongly advised to stop the DrivePool Service (and restart it or reboot when done). Also, I realise it is painfully slow for you (and I have no clue as to what would be the cause of that) but in my experience, DP is really a fire-and-forget program. I wouldn't care if it took a month so to say, it works just fine while it is balancing.

  8. ltmgtfy:

    :D

     

    Anyway, if you must, I would recommend to read that thread first, especially the 3rd post. Moreover, IMHO it'd be beteer to split the OS drive into two partitions and only add the non-OS partition to the Pool (again, only if it really benefits you to use OS HW in the Pool).

     

    Edit: I should read better, the post I referred to is not quite what you ask. In any case I think the answer remains the same: Yes you can, questionable whether you should.

    Also, the actual OS install (say windows, the Program Files and Users folders etc) will not then end up *in* the Pool.

  9. But then you could:

    1. Attach USB disk and add to the Pool;
    2. Remove one HDD from Pool (through GUI);
    3. Replace that HDD with a big one and add that new one.
    4. Remove USB HDD from Pool (through GUI);
    5. Depending on size and duplication, you can do the other three all at once or one by one.

  10. That it says your drive is full is strange and some screenshots might work.

    Have you set x2 duplication? In that case, two duplicates are stored on the 4 HDDs and no duplicate is stored on the same HDD. So you should be fine with the failure of one HDD. There is no reason, from that perspective, to use a 2x2 config (in fact, it becomes riskier in the sense that if you have 5TB of (unduplicated) data, DP can recover from 1x4 HDD to 1x3 HDD, using 10TB out of the 12TB remaining. In a 2x2 config and one drive failure, there is no recovery until you replace the faulty disk, which you would do anyway but until it arrives you would not have x2 duplication in full and I would consider switching off the machine until a new drive is available. I have a 2(Pools)x3(HDDs) config but ensure that each Pool can miss a drive.)

    That may change, depending on the backup software used, if you want to perform a backup as well and ensure only one duplicate is backed up. In the case, you might think of Hierarchical Pools. The search function should point you to some threads where I explained such setup.

×
×
  • Create New...