Jump to content
Division-M Community

KingofRome

Members
  • Content Count

    27
  • Joined

  • Last visited

  • Days Won

    2

KingofRome last won the day on December 27 2014

KingofRome had the most liked content!

About KingofRome

  • Rank
    Member

Profile Information

  • Gender
    Not Telling
  1. I've had this happen before when i was using a small SSD as a landing zone. My transfers were happening over the network so there might be a difference there but the same issue arose where the transfer was terminated as the landing zone couldn't be cleared fast enough while the transfers were happening and it reported the drive was full. Check your Balance settings for how often they are set to clear the landing zone (maybe set it to every 5 mins?). Otherwise best solution really is expand the landing zone to a bigger drive that can handle all the data your throwing at it at any given time. In reality the landing zone isn't useful if it needs to be cleared while copying files to it, as your speed is halved with the HDD being written to and reading other files to copy over to another drive happening at the same time.
  2. I know what you mean. I've had one of my MediaSonic enclosures fail and had a warranty swap already. And the stupid USB 3.0 thing dying under load i've definitely experienced where DB shows the drives as disconnected and the only way to recover was through a full system restart. So far though I've personally found that eSATA has been extremely stable and more than fast enough for the purpose of storage and access to the files, at least for my use as the computer is being used solely as a Gigabit NAS PC. Its primarily used to stream media through Plex. I've found on my setup that through the enclosure via port multiplier, i get about 250MB/S throughput and theres no way i can saturate that link via a gigabit network (Extra info: I have both the SI3132 and an on-board Marvell e-Sata controller chipset, the Marvel performs as above but the SI3132 does max out around 160MB/s so i think there are controller differences, the eSATA world is a total mess, so the card you mention doesn't sound bad at all. If i need to do any upgrades i might consider it though it would require a separate case and PSU.). I do have Smooth Streaming on for all files. Have tried in the past using DB with it off but i do find that with it on transfer rates are slightly slower but more consistent and overall faster. I haven't tried polling the drive via commands, but majority of my testing for polling and opening folders is occurring over the network. E.G: Over network and Local nearly identical in load time - Test is a folder 5 levels deep from main DB drive. 1434 items, 5.46GB, all .jpg files. 23 sec to just load entire folder before previews. To me this seems too long for the folder size, especially since the I/O is spread across 12 different drives, but let me know what you think, and totally appreciate your input.
  3. Wow, nice post and your set up is massive. I've noticed for folders with lots of files like pictures it simply takes forever to load, do you have folders that are similar or do you only have large video files? Could be a difference. Speed of each drive is more than enough and they are all connected either SATA 3 (via a Port Multiplier Mediasonic 4-Bay enclosure) or SATA 6 direct to the Mobo, so their should be enough bandwidth for them. I am going to double check though that the drives aren't spinning down even though I thought I set them to never spin down via windows settings. And now that you've said that, I honestly think having the NAS drive is more trouble than it worth attached to the pool then, because with it everything is much slower, as the NAS drive itself isn't exactly top tier.. Will follow up soon!
  4. Hey TheSmith, Was wondering if it were possible to utilize extra ram in a system to hold Windows index files of the pool via RAM caching. What I've noticed is that as soon as a NAS HDD is added to the pool, the whole pool slows down. I suspect this because it polls the NAS drive for files and therefore sometimes it leads to waiting a while for folders to load up as it finishes scanning all the drives. RAM caching index files of the drives or storing indexes somewhere else where it may be faster to read (the system drive, especially if it were an SSD) i could see definitely speeding this process up. Not even sure if its possible with DB, but i do know there is software that exists for RAM caching purposes that does a similar job (PrimoCache ect). As well, wondering if there was a way to use RAM caching more extensively than just with reading files via smooth stream but writing as well (ie PrimoCache). I have an HDD as a landing zone now, and i could use an SSD. However, if I used an SSD, it would wear it out so fast with transcoding and remuxing large movie and TV show files that with a large enough pool of RAM if might be better to go that route. A lot of my issues are with the fact that my DB pool is pretty massive (22TB, 12HDD's) and i'm seeing issues as it scales, so i know not everyone is going to seem them. But the opening folders lag, especially folders with lots of pictures can be brutal.
  5. 2nd this feature, mostly because I've had to pull 2 NAS HDD's within 2 days because of file corruption issues and lost a bunch of files.. Would be nice to scale duplication or some other form of drive backup, especially with bigger HDD pools, than only being able to have a single drive failure before issues and your data is unprotected during the rebuild period.
  6. Just loaded it today and everything seems to be going well.
  7. To me it looks like you added the drives, instead of merging them into an existing pool. Therefore the files exist on the drives themselves but your mount point of all the pooled drives looks empty. To access a pooled drive, you need to create a mount point... although from the looks of things you got like 8 mount points created for the same pool of drives. (L, M, N, O, R, Q, P, I... lol) Best bet, remove all the drives from the pool (the data you had on the drives before, outside the pool, is still safe, any data in the pool should be copied over to a temp drive somewhere. Then re-create the pool using the add drive and Merge feature. This will copy all the files of each drive into the pool. Don't know if you need to have one drive already in a pool to add the remainder of the drives in a merge. The convert features is good for trying out DB, but if you are planning on using it with data on the drives, merge is def better.. Also, your drives are REALLY full... should set Drive Bender do to some "even" balancing to help spread that data to your 2 empty drives.
  8. Wanted to let you guys know that i've installed the new beta and it works perfectly fine. Everything loads and is accessible. I just installed overtop of 2.0.4.0. Just throwing my 2-cents.
  9. I am having the same issue with both Health Check and Drive Balancing. If you try to start either manually, it doesn't work. Duplication Validation does work though.
  10. I'm have my mount point disconnect and not be able to reconnect.. even though there is drive activity going on... Its kind of serious as the only way to resolve is a complete system restart. Links to files below: Log files - https://dl.dropboxusercontent.com/u/4405769/DB%20Tests%20in%20Windows/Logs.zip Screenshot of activity occurring in DB consosle - https://dl.dropboxusercontent.com/u/4405769/DB%20Tests%20in%20Windows/Screen%20Shot%202014-05-15%20at%2010.40.00%20PM.png
  11. DB has a huge set of admin tasks to do after adding any single drive to a pool (under health check and balancing tasks.. you can see it in the real-time monitor).. it takes even longer if you have CRC verification on (which i recommend keeping on). I don't think you have to worry about losing data. I had this issue myself and system performance dies (even with the OS running off the SSD) as the computer has to poll the busy hard drives being analyzed by DB. Now imagine the time it takes to do 16.3 TB of slow-spinning hard-drives... With a migration of that size, it could take days. So i think you just got to be patient. Also this is common at least for me an easy to over-look. Double-check the task manager update frequency isn't set to pause, as i've noticed it automatically pauses itself sometimes instead of updating normally.
  12. Def sounds like drive bender will work without incident just by merging the drives into a pool. Though i recommend getting some kind of backup after that with that much data on the line. You also won't be able to do separate mount points (or "drives" as you would see them in windows) for the content because your content is mixed between all the drives. Instead, after you merge all the drives into a single mount point (of a single DB pool) and just make multiple folders and create separate SMB shares of each folder (so they'll appear separate on the network). You'll probably have to reorganize your data once its in DB after having all that data scattered among multiple drives.
  13. First thing. Is the HDD spinning down? because its separate from the OS and likely most often inactive it could be slowly spinning up when Plex accesses the database files and fails. this may not be the case as you say VLC isn't working either. 2nd thing. Why is your OS drive running off an E-sata? 3rd thing. So the way i understand it, the videos off the Drive bender pool cannot be played directly on the server with VLC (without Plex)? Are you only trying to play over the network?
  14. I opened a ticket to try and figure it out. Anthony and I discovered that it was due to errors on the individual hard drives. We had to mount the drives individually and run a chkdisk on each drive. After that it corrected the errors and i was able to delete the folder no problem! I did check the character limit and i know i'm under it. I didn't run an elevated command in the command window but i tried other things with administrator privileges and nothing worked. Totally a weird situation, but thanks for your input you guys!
  15. Hey all. I'm trying to delete a folder, it was an old backup folder i used to use when backing up my Pivos XIOS XBMC box and i was doing network backups to my DB Pool. I'm not using it anymore, so i tried to delete the folder itself, but windows keeps popping up saying the folder cannot be found after discovering over 60,000 files. I tried going into the folder and deleting the sub-folders and files there but then it comes up saying some files are missing. I have no idea how to get rid of the files and within the DB manager you don't have a file manager or something that might work differently. Looking for some input. Thank you!
×
×
  • Create New...