jbrodriguez

hi xxredxpandaxx, what is the log (/boot/logs/unbalance.log) showing ? Would you please reply in the support thread https://lime-technology.com/forum/index.php?topic=45352.105 ?

Hi, I don't get it, it just says stopped. Are you running the latest version ? I remember some 1.2.x version used to do that What's showing up in /boot/logs/unbalance.log ?

Ok, I see now I think including the cache disk shouldn't be too difficult. I will look into it. I'm still figuring out disk-to-disk transfers. Isn't a straight rsync a better fit for the job at hand? After all there are no free space calculations involved, just data transfer. 1) Consolidate Shares: This would read the split level of a desired share and then verify if it's being obeyed; if not, calculate probable scenarios to consolidate it or suggest the user to change it if it's not optimal. Are you thinking of something along the lines of consld8 (http://lime-technology.com/forum/index.php?topic=36201.0) or http://lime-technology.com/forum/index.php?topic=33689.0 ? It's interesting to see different ways people want to transfer data around in an unRAID array This is an interesting read as well: http://lime-technology.com/forum/index.php?topic=2689.0

Hi fonzie, No, unBALANCE doesn't work with the cache drive. I don't have one in my system, so I can't really help out, but a suggestion would be to use rsync. It's not that difficult in the end. Check the thread for rfs to xfs data transfer, it should be helpful.

Rebel, what happens when you click on the green Open Web UI link ?

I've published v1.3.4, which adds progress stats when moving folders. It has a folder level granularity (not file level), so increments should be in big chunks, and spaced in time between them. But it gives a ball park estimate of where the operation is sitting at. I also took a look at what smashingtool reported, with the app stopping and all. It's very difficult to replicate, what I've read is that sometimes rsync hangs, but there are no specific reasons. If someone comes across this issue (app stopping/rsync hanging), please provide me with the output of the following commands $ ls -l /proc/$(pidof rsync)/fd/* $ ps aux | grep diskmv $ ps aux | grep find Additionally, to move on with the process, I'd do the following: 1 - Get the last diskmv command that's being executed $ ps aux | grep diskmv It should output something like ... /bin/bash /usr/local/emhttp/plugins/unbalance/diskmv -f films/bluray/Movie/ /mnt/disk1 /mnt/disk2 2 - Stop unBALANCE from the plugin settings page 3 - Execute the diskmv command manually (minus the /bin/bash) $ /usr/local/emhttp/plugins/unbalance/diskmv -f films/bluray/Movie/ /mnt/disk1 /mnt/disk2 This should sync the folder and the data should be in a reasonable state. 4 - Start unBALANCE again, calculate, move

Actually, there was a bug with progress messages, I've fixed it in 1.3.2

Like CRD mentioned, it can take a while if there's lot of small files. But, it's showing the progress bar, next to the 'started' label right ?

I've published v1.3.1, with the following changes: - Use ip address instead of hostname when opening app window - Improve UI feedback when an operation is in progress (browser is being reopened) - Set a 1000 lines buffer for the console (older messages will be removed) Let me know if you find any issues. I have to think some more about stopping the operation. With regards to the copy functionality, I'm about 95% sure I won't do it

... ok sorry so it looks like it is still running but the website can't tell so it allows you to run it again?? can the header be made to reflect if it is truly running? and lock out the start like a parity check? sorry for the trouble but i am guessing that i must have started it 3 times now on top of each other prior to knotting that it was running. Like trurl mentioned, it won't let you click on the button if it's already running, do you see something different ?

You're right, I've seen that too. I need to fix it.

Thanks for the comments lishpy ! I'll take a look at the link opening code, to see if I can use ip address rather than hostname (probably, that's what's happening to huladaddy too). I'm a bit tied up until the end of next week, but I'll look into it. no idea what the webUI page is, but if you just have it as another .page file, then reference it as Settings/webUI and dynamix will handle everything for you... (turn on help, and look at the link for CA's manual, and look at caHelp.page Squid, in the settings page, when the app is started, there's a link to the app's user interface ... it's opens a new window/tab with the server's hostname, rather than ip address, I'm not sure if there's an env variable that holds the ip address can't you just do <a href="/settings/Unbalance.page" target="_blank"> and have unraid figure it out for itself? IIRC the browser will correctly go to either the IP or to the hostname depending upon how the originating page was loaded. I'll look tonight. don't remember why I was trying env variables, but your suggestion seems saner, will check it too

I always use the latest OSX version and never had any problems with browsing unRAID shares. Does it work fine from the command line ? I always disable writing .DS_xxxx files to network drives ... not sure if that might help, but worth a shot.

For me ... unRAID > other solutions, because: 1) Real time parity protection: I don't use a cache pool, I can live with network performance, while my mind is at ease 2) Mix and match disk sizes: Inevitably, disks get bigger, so I get two bigger sized disk (one for parity, one for data) and keep growing my array. 3) Grow your disk pool economically, compared to other (ZFS I'm looking at your vdevs! But I have to say, ZFS has some awesome use cases and is awesome itself !) 4) Ease of use: even if personally I find preclear to be a PITA, after that you're in the clear. I tried snapraid once and I had to keep messing with some block size, which never worked for me. 5) The Forum: very knowledgeable people willing to help guys out. In some other forums, I found sarcastic comments and basically go f... yourself if you don't know how this is supposed to work I've had a couple motherboards die on me, never lost any data with unRAID ... I'll quote Hernandito .. "someone would have to pry my unRAID from my cold dead hands..."

Thanks, It's great that you made that, I'll give it a shot. My only other question now is whether it's possible to pause or stop unBALANCE after you start it. You can't stop it right now, except for killing the unbalance process and the rsync processes from the command line. unBALANCE will go down gracefully as soon as it is interrupted, but rsync will go down hard. Having said that, I think it's possible to add stopping the plugin in the middle of a move operation. I need to look at some scenarios to see how it might work. I don't know if the stock mover can be stopped as well ... it would help to check out how they did it, if they did

hi smashingtool, Do you see anything out of the ordinary in either unraid's or unbalance's logs ? The heart of the operation is an rsync command, any error it might come up with should be written to the log. But rsync can also slow down/pause with a failing disk. Let me know if you find any other evidence of what might have happened.

No point in using remove-source-files if you are planning on formatting the disk anyway. Copying vs moving is much faster. I have: Parity Disk1 - XFS Disk2 - RFS Disk3 - RFS Disk4 - RFS Disk5 - XFS (New drive, no user shares) I've just done Disk1, my plan is to use Disk 5 as a placeholder while I empty RFS drives followed by reformatting. So I'm using unBalance to move files from the RFS disk to my placeholder. Then I reformat it to XFS. Since there are no user shares on the placeholder, I can't use unBalance. I have a specific requirement for having a certain type of file/shares on certain disks, otherwise I wouldn't be moving them back. Should I use the diskmv script diskmv -f "" disk1 disk2 instead or maybe just mv mv /mnt/disk1/ /mnt/disk2 ? Hi CRD, If you're copying disk to disk, which means you have available space on the target disk, I would suggest doing a straight rsync copy .... there's a thread somewhere in the forums about this same topic (rfs to xfs transfer)

Hi Unqualified Spectator, thanks for your kind words ! I really haven't thought about a use case where you're just copying files, not moving them. I'd like to share how I came up with the idea for this app/plugin The thing is .... I regularly deleted low quality content that I replaced with higher quality stuff or even deleted some content that I found uninteresting. This made some disks in my array to get some free space, which was wasted. But manually calculating which folders I could move from other disks to use that space was not scalable. I think I half did it one time And so ... unBALANCE was born ! I'll take a look at the idea over the weekend, but it does seem like a different concept.

Thanks for the comments lishpy ! I'll take a look at the link opening code, to see if I can use ip address rather than hostname (probably, that's what's happening to huladaddy too). I'm a bit tied up until the end of next week, but I'll look into it. no idea what the webUI page is, but if you just have it as another .page file, then reference it as Settings/webUI and dynamix will handle everything for you... (turn on help, and look at the link for CA's manual, and look at caHelp.page Squid, in the settings page, when the app is started, there's a link to the app's user interface ... it's opens a new window/tab with the server's hostname, rather than ip address, I'm not sure if there's an env variable that holds the ip address

Thanks for the comments lishpy ! I'll take a look at the link opening code, to see if I can use ip address rather than hostname (probably, that's what's happening to huladaddy too). I'm a bit tied up until the end of next week, but I'll look into it.

I have the same problem -- blank page. Is there another way of getting unBalance to run? I am currently preclearing four disks and can't restart unraid. Hi huladaddy, not sure why you come up with a blank page. If the plugin is running (it shows the green Open Web ui link), you should be able to access it via either http://ip:6237/ or http://hostname:6237/, following whatever addressing scheme you have in your network. Does the log show that it's actually running (/boot/logs/unbalance.log) ?

Yes, it copies then deletes. diskmv has a clobber option (-c "clobber, duplicates on destination will be overwritten"), but I'm not sure how sensible it would be to use it in the general case.

Not sure why it doesn't pick up the latest version, unless you reboot or start/stop the plugin. I'll check the install/upgrade code.

It isn't showing up in the Settings page ?

I've published v1.3.0. Features/Bug Fixes: It allows you to choose the reserved minimum space (in either Mb, Gb or a percentage), but still keeps a minimum hard limit of 450Mb (per kizer) It also has a better check to prevent adding a subfolder of an existing folder and finally the plugin is disabled by default (per Ryland)