bmfrosty Posted January 14, 2018 Share Posted January 14, 2018 I hit the upgrade, but maybe the upgrade didn't actually upgrade. Will try again. Quote Link to comment
bmfrosty Posted January 14, 2018 Share Posted January 14, 2018 I stopped and started unBalance, and it was the newer version. It worked on my previously problematic sample. Thanks for the fix! Quote Link to comment
deaerator Posted January 15, 2018 Share Posted January 15, 2018 After the update to 6.4 and setting up ssl certificate, I can no longer connect to Unbalance Web gui. Quote Link to comment
jbrodriguez Posted January 16, 2018 Author Share Posted January 16, 2018 On 1/12/2018 at 6:13 PM, ashman70 said: is there a way to just select all the data on say disk 1 without having to manually check Currently, not Quote Link to comment
jbrodriguez Posted January 16, 2018 Author Share Posted January 16, 2018 On 1/13/2018 at 11:11 AM, RGauld said: I no longer receive that message when I boot unRAID now. Thanks for the confirmation and the kind words ! Quote Link to comment
jbrodriguez Posted January 16, 2018 Author Share Posted January 16, 2018 On 1/13/2018 at 9:00 PM, woble said: Easy to fix as follows. Thanks woble ! I'll check out how this works. Quote Link to comment
jbrodriguez Posted January 16, 2018 Author Share Posted January 16, 2018 On 1/14/2018 at 4:40 PM, bmfrosty said: It worked on my previously problematic sample. Awesome ! Quote Link to comment
jbrodriguez Posted January 16, 2018 Author Share Posted January 16, 2018 20 hours ago, deaerator said: After the update to 6.4 and setting up ssl certificate, I can no longer connect to Unbalance Web gui. Yes, it works over http only right now. If you want to access the UI, just connect to http:// instead of https:// It's in my to-do list to get it working on https. Quote Link to comment
JustinChase Posted January 19, 2018 Share Posted January 19, 2018 a search didn't reveal anything. Why have the 450MB reserve requirement? I want to completely fill my full disks, so they can't be written to. They are just for storage, so there's no point in me leaving space and having unRAID try to manage such a small amount of free space. Sadly, I didn't realize this issue until I'd filled most disks, so most have only this amount of free space, and many of my files are larger than this, so it will not be easy to fill up these disks without much work of me finding small files and manually moving them. Is there any way around this limit? the program says it's hard limited, so I guess not, but thought I'd ask. Quote Link to comment
jbrodriguez Posted January 19, 2018 Author Share Posted January 19, 2018 Hi JustinChase, The main reason for the hard-limit is the issue of file size vs sector size. If you have a 1 byte sized file it still occupies a full disk sector (it's still 4k these days right?) and these 4k bytes count towards the total available free space. If you have A LOT of small sized files, you could have many sectors with "unused" free space. So, I settled on a 450Mb limit, to account for this unused disk sector space. Later on, via this forum, I also came to know that ReiserFS doesn't like mostly filled disks that much (unBALANCE was born when RFS was the norm). I could have increased the hard limit, but since the min free space is configurable, I decided to leave it be. I understand that XFS is better behaved with regards to fully populated disks, but people in the forum with knowledge on the matter still recommend not filling up a disk. Hope this makes sense. Quote Link to comment
GumbyAdmin Posted January 21, 2018 Share Posted January 21, 2018 Greetings, I have attempting to use the Scatter/Move twice, and both times it fails out at the end of the 1st Source copy. I've tried moving to 2 separate disks, both with plenty of free space afterwards, and it still fails. The unbalance.log file only shows exit status 23: Partial transfer due to error, it doesn't say what file it ended on. I was able to use the rsync command through ssh without issue, other than it leaves the source files. Quote Link to comment
RGauld Posted January 21, 2018 Share Posted January 21, 2018 On 1/19/2018 at 11:28 AM, jbrodriguez said: I could have increased the hard limit, but since the min free space is configurable, I decided to leave it be. I understand that XFS is better behaved with regards to fully populated disks, but people in the forum with knowledge on the matter still recommend not filling up a disk. What would you suggest filling the available hard drives to? I have my array set to fill approx. 85% of my drives.... Quote Link to comment
JustinChase Posted January 21, 2018 Share Posted January 21, 2018 On 1/19/2018 at 1:28 PM, jbrodriguez said: Hi JustinChase, ... Hope this makes sense. It does. Thank you again for sharing your work with us. It's been very helpful! Quote Link to comment
jademonkee Posted January 23, 2018 Share Posted January 23, 2018 Ah balls. After a failed transfer "exit status 23: Partial transfer due to error" I went to re-run the move and accidentally selected the incorrect destination disk. Rather than waiting an hour and a half for it to finish, and then having to figure out how best to distribute the files to the correct disk, is there any way to cancel the operation? I can't see one. Thanks. Quote Link to comment
jademonkee Posted January 23, 2018 Share Posted January 23, 2018 1 minute ago, Fireball3 said: Hah. I searched for 'cancel' before I posted, but it didn't return that result (assuming because it was written as 'cancelling'). And it's located on the previous page, too! So embarrassing. Thanks for the info. Quote Link to comment
Cessquill Posted January 23, 2018 Share Posted January 23, 2018 3 minutes ago, jademonkee said: Hah. I searched for 'cancel' before I posted, but it didn't return that result (assuming because it was written as 'cancelling'). And it's located on the previous page, too! So embarrassing. Thanks for the info. The suggested option worked well for me 1 Quote Link to comment
Fireball3 Posted January 23, 2018 Share Posted January 23, 2018 With regard to the error 23 I wrote my experience here: I haven't used the plugin since then, so I can't say if those issues have been resolved yet. 1 Quote Link to comment
mrow Posted February 1, 2018 Share Posted February 1, 2018 (edited) Can you give us the ability in the settings to set all the flags on our own? The -X flag to copy extended attributes is causing a ReiserFS error and thus causing a move to fail. I tried using my own flags but realized, after it failed again and looking at the logs, that putting custom flags in the settings page just adds those flags in addition to your standard flags. Please give an option for us to use only our custom flags. It’s not a huge deal. I’m just running the rsync operation from the command line for now. Edited February 1, 2018 by mrow Quote Link to comment
jbrodriguez Posted February 1, 2018 Author Share Posted February 1, 2018 16 hours ago, mrow said: Please give an option for us to use only our custom flags. I used to have fully customizable flags, but the app really needs to "guard" itself especially with -a: since it will be replaced with a -c, for checksum/validation -R: because it needs to have the working directory set to the source disk at least But I can see that -X isn't really a core flag and can be left on as customizable default (removable). I'll do this. In any case, I'm looking into an alternative rsync invocation that I believe will bring execution parity with a command line invocation. I'm still doing some research and testing, but it looks promising. Quote Link to comment
Joseph Posted February 3, 2018 Share Posted February 3, 2018 On 1/8/2016 at 3:40 PM, jbrodriguez said: Description unBALANCE helps you manage space in your unRAID array, via two different operating modes: Scatter Transfer data out of a disk, into one or more disks Gather Consolidate data from a user share into a single disk Hey jbrodriguez, If you are open to thoughts/ideas on adding features, please consider adding another mode which will auto-magically 'defrag' (for lack of a better term) all the folders based on the starting depth. So it would be like using Gather Mode, but unBalance will take the time to reorder things across all the disks. For example, if a share is spread across 5 disks, it would find where the bulk of the data is and move the data from other drives to that drive. If there isn't enough space, it will figure out the next best place for it, or perhaps it would move other folders around to achieve the best 'defrag' (again, not the right word) option for the share at the starting depth the user selected. Hope this makes sense, it does in my head. Quote Link to comment
trurl Posted February 3, 2018 Share Posted February 3, 2018 1 hour ago, Joseph said: Hey jbrodriguez, If you are open to thoughts/ideas on adding features, please consider adding another mode which will auto-magically 'defrag' (for lack of a better term) all the folders based on the starting depth. So it would be like using Gather Mode, but unBalance will take the time to reorder things across all the disks. For example, if a share is spread across 5 disks, it would find where the bulk of the data is and move the data from other drives to that drive. If there isn't enough space, it will figure out the next best place for it, or perhaps it would move other folders around to achieve the best 'defrag' (again, not the right word) option for the share at the starting depth the user selected. Hope this makes sense, it does in my head. Sounds like a sort of Split Level after the fact. Do you know how to use Split Level to prevent this "fragmentation" in the first place? Quote Link to comment
Joseph Posted February 3, 2018 Share Posted February 3, 2018 12 minutes ago, trurl said: Sounds like a sort of Split Level after the fact. Do you know how to use Split Level to prevent this "fragmentation" in the first place? I'm somewhat familiar with it, but perhaps not enough. As I understand, even with Split level, folders/files will become fragmented over time as things are deleted and space is freed. Based on the size of new data being moved to the array, there may not be enough room to store everything on a single disk; so at some point when the array starts to get full, Split level will then split it onto 2 or more disks. You are correct that what I was proposing to jbrodriguez is like split level after the fact for situations such as this... or if someone didn't use split level in the first place (yours truly) and would like to get as much things 'unscrambled' as possible. The idea here is it would intelligently 'look' at the folder and depth selected by the user, and move even other data around as necessary to achieve the best fit possible for everything by minimizing splits as much as possible. Quote Link to comment
trurl Posted February 3, 2018 Share Posted February 3, 2018 1 hour ago, Joseph said: I'm somewhat familiar with it, but perhaps not enough. As I understand, even with Split level, folders/files will become fragmented over time as things are deleted and space is freed. Based on the size of new data being moved to the array, there may not be enough room to store everything on a single disk; so at some point when the array starts to get full, Split level will then split it onto 2 or more disks. You are correct that what I was proposing to jbrodriguez is like split level after the fact for situations such as this... or if someone didn't use split level in the first place (yours truly) and would like to get as much things 'unscrambled' as possible. The idea here is it would intelligently 'look' at the folder and depth selected by the user, and move even other data around as necessary to achieve the best fit possible for everything by minimizing splits as much as possible. Split Level takes precedence over other User Share settings, so I think even if a disk gets too full it is still going to try to keep things together according to Split Level and just let the write fail if there isn't room. I think another way to accomplish this "after the fact" at the command line (or mc) would be to get your Split Level like you want it, then move things from user0 to cache and let mover put it back. Quote Link to comment
Joseph Posted February 3, 2018 Share Posted February 3, 2018 23 minutes ago, trurl said: Split Level takes precedence over other User Share settings, so I think even if a disk gets too full it is still going to try to keep things together according to Split Level and just let the write fail if there isn't room. I'll have to re-examine split level. Mostly, I need to rethink my shares strategy instead of dumping everything into one user share. 27 minutes ago, trurl said: I think another way to accomplish this "after the fact" at the command line (or mc) would be to get your Split Level like you want it, then move things from user0 to cache and let mover put it back. There's more things scrambled than not, but there are several ways to go about 'defraging' them for sure. It would just be nice to have a utility that could do that so it wasn't a manual process. Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.