Jump to content

Parallax Abstraction

Members
  • Posts

    13
  • Joined

  • Last visited

  • Days Won

    3

Reputation Activity

  1. Like
    Parallax Abstraction got a reaction from Christopher (Drashna) in Macrium Reflect Reporting Out of Disk Space Mid-Way Through Backup   
    OK, fair enough. I turned off duplication on the backup image folders as I don't need that anyway and it allowed me to complete an initial image without issue. Should be good now as the incremental will be much smaller. Thanks!
  2. Like
    Parallax Abstraction reacted to Christopher (Drashna) in Macrium Reflect Reporting Out of Disk Space Mid-Way Through Backup   
    StableBit DrivePool doesn't split the files up itself.  It stores the actual files on the underlying disks. So you can't have files larger than the disks in the pool, unless they're broken up into chunks. 
  3. Like
    Parallax Abstraction reacted to Parallax Abstraction in Macrium Reflect Reporting Out of Disk Space Mid-Way Through Backup   
    I might be wrong but I thought DrivePool was able to use striping to take ultra big files like this and spread them across multiple disks when writing. Indeed, if I look at my MediaSonic ProBox while running the backup, I can see lights for two different drives flashing. However, if it does do that, maybe it does it after individual files are written, I'm not sure. Oddly though, the WD Black drive in my pool has 1.97TB of free space and that's definitely more than the backup will be. Unless the balancing plugins are causing it to be written elsewhere first. Truth be told, it's not really a big deal if I have Reflect put a hard cap on the file size so I'll try to run a backup that way and see if DrivePool can handle it. I'll set the limit to 500GB per file and report the results. If the devs have any insights though, that's be awesome.
     
    EDIT: So it turns out, if you're using file splitting in Reflect, you can't use incremental backups. That's a major issue for me as my custom incremental retention settings are one of the main reasons I use this product.  I'm going to try something else though. Rather than having the entirety of the DrivePool be duplicated, I'm going to switch to folder level duplication and have it exclude the backup folder, which I definitely don't need duplicated anyway. I'll see if that helps.
  4. Like
    Parallax Abstraction reacted to Spider99 in Macrium Reflect Reporting Out of Disk Space Mid-Way Through Backup   
    Hi i'm new to this as well - but the thought that comes to mind is that the backup is a single file and you dont have enough space on two drives to hold the single file. So i suspect the out of disk space warning is when one of the drives with very low space fills up as MR probably does not reserve the full file size initially on the system(if it did then the backup would probably fail almost immediately) because it does not know how big the backup is going to be and just incrementally expands the backup file as it goes later it will know when the process nears completion. DP therefore just decides to pick a disk and wait for the process to finish hence why you get variable completions of 20% and 40% as the "primary" or "secondary" DP disk fills up.
     
    Drive 6 looks like it has enough space but the others look to have a lot less space.
     
    If you can chop it up into more manageable pieces then I would expect it to work. Or just do an image of your system drive and do a file backup of the rest to keep the biggest file size down to something more manageable
  5. Like
    Parallax Abstraction reacted to Parallax Abstraction in Macrium Reflect Reporting Out of Disk Space Mid-Way Through Backup   
    Hey all! I'm still on my trial of DrivePool but am overall liking it quite a bit. However, I've run into one major problem I don't know how to solve and I'm hoping someone here knows a solution.
     
    I use the paid version of Macrium Reflect to backup my entire gaming PC. It writes a main image that's a full backup (it usually comes out to about 1-1.3TB after compression) and then writes up for 4 incremental backups after that, then rotates those out, always keeping no more than 5 backups. My DrivePool lives on my home server under my desk and I run the Macrium Reflect backup set to a share I have on that server. When I tried to start my backup, my pool looked like the attached screenshot. About 4TB of space left between a few drives. When I execute my Macrium Reflect script, it also reports the same amount of space remaining.
     
    However, multiple times today I've tried to run the full backup and it's failed between 20% and 40%. According to the Reflect logs, it's because it...ran out of space. Even without compression, the total backup from my gaming PC is less than 2TB, it most certainly is not running out of space. As you can see from the screenshot, I do have duplication turned on but even with that taken into account, there's still more than enough space available.
     
    I'm not sure why this would be happening or what to do about it. I can engage Macrium support as well but I figured someone here might have a better idea. I can set Reflect to split the backup into multiple smaller files and may try that but I'm not sure why that would be necessary. I'm not sure why writing one large file over the network would keep failing this way.
     
    Anyone have an idea why this might be happening and if there's some kind of solution? Thanks all!

  6. Like
    Parallax Abstraction got a reaction from Christopher (Drashna) in Tried FlexRAID and Frustrated - Will DrivePool work for me?   
    Duly noted! Even if I don't use CloudDrive, for $5 why not.
  7. Like
    Parallax Abstraction reacted to Parallax Abstraction in Tried FlexRAID and Frustrated - Will DrivePool work for me?   
    Duly noted! Even if I don't use CloudDrive, for $5 why not.
  8. Like
    Parallax Abstraction reacted to Christopher (Drashna) in Tried FlexRAID and Frustrated - Will DrivePool work for me?   
    Well, glad to hear that it eventually removed the drive! 
     
    And glad that the software is working very well for you! 
     
    And just FYI, the StableBit Bundle is the best deal.  You basically get StableBit DrivePool and StableBit Scanner at the normal (discounted) price, and StableBit CloudDrive for $5. 
     
     
    If you need anything else or have any more questions, don't hesitate to ask!
  9. Like
    Parallax Abstraction reacted to Parallax Abstraction in Tried FlexRAID and Frustrated - Will DrivePool work for me?   
    Happy to say that it chugged and appeared frozen for a while but DrivePool did pull out the busted Seagate and I didn't lose anything! It's out of the pool now and when my new 8-bay MediaSonic enclosure arrives, I'll be putting in two other drives, which will give me the means to use full duplication. I am super impressed with DrivePool. When I get paid next, I'll be buying licenses for both it and Scanner and CloudDrive seems neat too and I'll be trying that out soon as well. Great work guys, I should have gone with this from the start!
  10. Like
    Parallax Abstraction reacted to Parallax Abstraction in Tried FlexRAID and Frustrated - Will DrivePool work for me?   
    Thanks for the answers! So I was losing my mind with FlexRAID and got it out of my life last weekend. I'll never touch that product ever again. I decided to go ahead with a DrivePool trial and I gotta' say, I'm pretty damn impressed so far.
     
    There's a real test going on as I type though. So I have 4 WD 3TB Red drives and also added in a 4TB and 3TB Seagate drive I had from a previous Windows Storage Spaces pool. I was also running a trial of StableBit Scanner. Well, this morning, Scanner informed me that the 3TB Seagate is starting to fail. I'm not surprised, it's Seagate and I got it for free originally. I've I didn't have any duplication on yet but was seriously just about to enable it when this happened. Yeah, this conversion's been a nightmare. I've asked DrivePool to remove the Seagate (which has a lot less on it than the other volumes) and it's trying but it looks like that drive is failing very fast and it's really struggling. It says "Removing drive..." but also shows no read or write activity on the pool so I don't know if it's actually working. If it can manage to actually get the data off and let me passively remove it, that'll be pretty amazing.
     
    I think this combined with CrashPlan will be a great fit for my setup. My plan is to expand the pool enough that I can do full duplication so the CrashPlan issue that the other user had shouldn't really be a factor for me. We'll see how things go with this busted drive but I'll very likely be purchasing DrivePool and Scanner. I'm so glad I never got to the point of actually spending money on FlexRAID.
  11. Like
    Parallax Abstraction reacted to Christopher (Drashna) in Tried FlexRAID and Frustrated - Will DrivePool work for me?   
    Ugh, CrashPlan. Why do you have to use Java as the base?! 
    But yeah.....
     
     
    As for healing, no. Not currently. Implementation something like that is very complex, and prone to issues. However, we do flag mismatched files, when they are detected. 
     
     
    Yes. If a drive fails, .... well, it depends on how. If it no longer shows up in the system (or you pull it), it is flagged as missing and the pool is set as read only (to prevent potential corruption/mismatched data).  Once you've removed the missing disk, it checks a few things, reduplicates data as needed and rebalances the data as needed. 
    It immediately continues to work, though. 
    However, if you have StableBit Scanner installed on the system as well, it may detect issues before the drive outright fails. And in this case, it may trigger the software to evacuate the contents of the problem drive, hopefully before it becomes an issue.  Yes.  Well, rather, since each 3TB drive would show up as 2.74 TBs, you'll have roughly 10.8TBs of usable space. 
    However, this doesn't account for duplication.  This just shows the raw capacity.  And duplicated data will take up twice the space. So if you're duplicating everything, you'll have roughly 5.4TB of usable space.  Yes.  It shows up as a normal NTFS volume in almost all regards (I say almost, because we don't support VSS snapshots). CrashPlan should work on the pool properly.  It depends on how you're backing up the data. If you're backing up the pooled disks and not the pool, then yes, it definitely becomes more of an issue, because of the balancing.  
    However, I'm not really well versed in CrashPlan, as I haven't used it much.  Your best bet for answers is with the people that have used it more extensively.  Sorry. 
      And for all of the questions, not a problem!
  12. Like
    Parallax Abstraction reacted to Parallax Abstraction in Tried FlexRAID and Frustrated - Will DrivePool work for me?   
    Hey all. Forgive me if this post is long but I'm having a hard time making a decision here.
     
    So I have about 9TB of data that I am keeping on a small Windows 10 home server. It's mostly media, video projects and emulation stuff (I run a retro gaming YouTube channel and have tons of emulation sets.) It's about 485,000 files in total. Until recently, it was all kept on a Windows Storage Spaces JBOD volume through a 4-bay MediaSonic ProBox that was connected directly to my main PC. Problem is, because CrashPlan has such a horribly written client, backing all this up to the cloud was taking over 3GB of memory at any given time. So I built this home server and moved stuff there. The problem was, some of the data was still residing on my desktop PC and again because of CrashPlan's crap client, it was constantly rescanning terabytes of data off my main machine over the network which also was no good. So I decided to move everything to my home server.
     
    I bought 4 new 3TB WD Red drives and decided to try FlexRAID RAID-F. I copied everything off the old Storage Spaces volume to a bunch of other drives I had lying around, installed the new drives and created a FlexRAID volume. 3 data drives, 1 parity drive. This was way more effort than it should have been as FlexRAID's UI isn't great, the documentation is terrible (tons of pages on their wiki are outdated), their forums are useless and the developer won't even talk to you unless you pay a ridiculous amount of money. Still, I got it working. Then I spent the better part of 2 days copying all the data back. Then I discovered that because of the way CrashPlan reads the file system, it is not capable of doing real-time backups with FlexRAID RAID-F and can only discover changes when everything is rescanned. Because of the size and number of files in my backup, this takes literally hours each time. That's no good, I pay for CrashPlan in part for real-time backup. I'm fed up with FlexRAID and am ready to copy everything back off it again and dump it before the trial ends.
     
    I've been looking at alternatives and am intrigued by DrivePool. What you seem to offer isn't RAID per ce, it's storage pooling but with the option to have some or all of the data duplicated across multiple drives to protect against failure. As I understand it, this can't heal itself from a failure like a RAID can but it is possible to have a drive die without losing the whole pool. Truth be told, I think having a full RAID plus CrashPlan is probably overdoing it for my scenario. I have 12TB of storage available with this pool of Red drives but right now, only have access to 9TB of it as I have 1 drive being used for parity. If I convert the whole thing into DrivePool, I will have an extra 3TB I can use to duplicate the most important stuff, while entrusting the rest to CrashPlan. Hopefully that explains what I'm looking for.
     
    So, before I take the plunge and copy all this data twice yet again, here's my questions for confirmation:
    If a drive fails, I can remove/replace that drive without it taking out the rest of the pool? If I pool all 4 of these 3TB drives together, I'll get a combined pool of roughly 12TB? Will DrivePool appear as a normal NTFS volume so that CrashPlan can back it up in real-time? FlexRAID RAID-F does not but their T-RAID option apparently does. I read a recent thread where someone talked about it being a nightmare to restore FROM CrashPlan after a drive failure, largely because of CrashPlan's crap client again. Should I run into a failure and it's with unduplicated data, is this what I can expect to deal with? It's not a deal breaker, I just want to know. I hate CrashPlan's client but unfortunately, they're the only truly unlimited option I have available. Again, sorry for the long post but after wasting so many hours on FlexRAID, I really want to make sure what I choose next will do the job. As I said, I think with CrashPlan, I don't necessarily need full RAID with this data. But I want to make sure I'm not missing anything. Thank you very much!
×
×
  • Create New...