Unionfilesystem Plugin

Okay, thanks - I'm running the update check now. Will let you know what happens.

multiplex: to combine multiple things to one.

If a single drive is filled by a single file there isn't anything mergerfs can do besides split the file up... which it won't do because the point is to simply merge drives together as is. Not to build a filesystem on top of them. The policies dictate which drive is selected for creating a file but that has nothing to do with writes. The only control over writes is 'moveonenospc' which in the docs should be clear on.

You need to read the section about policies and path preserving and what moveonenospc does.

You should probably not be using a path preserving policy and turn on moveonenospc.

I installed the backports kernel, rebooted, and ran the update. I didn't see any packages related to fuse (lots of firmware stuff), but I installed everything it presented to me anyway. Still having the same problem with copying additional files to the pool. Should I try reinstalling the fuse 2.9.3 package?

Trapexit,
Thanks - I'm not copying a single file. I'm copying thousands of files in hundreds of folders, with one top level folder common to all. I understand that a file cannot span across multiple drives, but a folder (path) should be able to span across multiple drives, right?
Best,
Phil

And each of those files is less than 8GB. My smallest drive is 2TB...so shouldn't be having a problem with spanning any file.

Sorry...I really appreciate your help.

What is your configuration? Are you using moveonenospc? What's your minfreespace? What policy are you using? What size are the files? Are they bigger than minfreespace?

Trapexit
I am using OMV 2.2.4 (stoneburner) with snapraid installed. I have installed the union filesystems plugin, and created a volume using mergerfs with epmfs policy. Minfreespace is set to 8GB. The files are all much smaller than 8GB, so I should never run into that issue (I had that issue with unRAID and learned from those trying days). I don't know if I am using moveonenospc or not...I am not fluent in command line, so I only use what is available in the GUI for configuration. I do not remember configuring moveonenospc nor do I know if it can be configured in the GUI. Any guidance you might have would be greatly appreciated. ryecoaaron is working on updating the backport, so it might not be a problem with mergerfs at all, but rather fuse. If you'll let me know what logs or info you need, I'll be happy to pull it up for you. Right now, I have uninstalled the unionfilesystems plugin while waiting on the back ports update. That way, the plugin will install with the most recent packages.
Best,
Phil

It's not a problem with fuse or mergerfs but your setup but without understanding **exactly** what you're doing and how it's setup we can't help.

Are you intentionally using path preserving? if you don't care what drive the files end up on you should not use a path preserving policy.

Trapexit,
I'm reinstalling the plugin right now (haven't installed any back port update for fuse) to test a non-path preserving policy. Perhaps that was the problem.

What else do I need to tell you to help you understand my setup? I'm not trying to be obtuse.
Best,
Phil

Trapexit,
I reinstalled the plugin, and chose mfs policy. The files are copying again.
Thank you.
So, if I'm understanding this correctly, preserving the path prevents any usage beyond one drive?
Best,
Phil

All the settings for mergerfs and a simple example of the problem with all information about drives, capacity, file size, etc. Everything related to what's going on.

The policies use the drive size at create time to determine where something goes. If I don't know what that information is I can't comment on why it's behaving a certain way.

Here ya go. Curious - I'm using a USB flash drive, with the flash memory plugin installed and I modified the fstab file as directed...but the changes appear not to have taken...

Zitat

# /etc/fstab: static file system information.

#

# Use 'blkid' to print the universally unique identifier for a

# device; this may be used with UUID= as a more robust way to name devices

# that works even if disks are added and removed. See fstab(5).

#

# <file system> <mount point> <type> <options> <dump> <pass>

# / was on /dev/sdf1 during installation

UUID=6594ce43-24ac-49c6-98a0-6eee47323d26 / ext4 errors=remount-ro 0 1

# swap was on /dev/sdf5 during installation

UUID=57c1957b-d96a-43a2-863a-68f76573c9ed none swap sw 0 0

/dev/sde1 /media/usb0 auto rw,user,noauto 0 0

/dev/sdf1 /media/usb1 auto rw,user,noauto 0 0

/dev/sdf2 /media/usb2 auto rw,user,noauto 0 0

/dev/sdf5 /media/usb3 auto rw,user,noauto 0 0

# >>> [openmediavault]

UUID=82989895-0f34-4655-95d4-46bcaf953fd6 /media/82989895-0f34-4655-95d4-46bcaf953fd6 ext4 defaults,nofail,user_xattr,noexec,usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0,acl 0 2

UUID=eec8ff77-1dc9-41a8-9e7b-f9ef53a5bf63 /media/eec8ff77-1dc9-41a8-9e7b-f9ef53a5bf63 ext4 defaults,nofail,user_xattr,noexec,usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0,acl 0 2

UUID=d76b7e96-fced-4edc-a32a-cbb20968db87 /media/d76b7e96-fced-4edc-a32a-cbb20968db87 ext4 defaults,nofail,user_xattr,noexec,usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0,acl 0 2

UUID=d7656b4c-0bac-4db8-825b-427a09d90b47 /media/d7656b4c-0bac-4db8-825b-427a09d90b47 ext4 defaults,nofail,user_xattr,noexec,usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0,acl 0 2

UUID=3b4c969a-9010-4883-a141-1758ddc8fa3c /media/3b4c969a-9010-4883-a141-1758ddc8fa3c ext4 defaults,nofail,user_xattr,noexec,usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0,acl 0 2

UUID=8060a38c-45d3-4571-855d-3830f71cb846 /media/8060a38c-45d3-4571-855d-3830f71cb846 ext4 defaults,nofail,user_xattr,noexec,usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0,acl 0 2

UUID=fd9cf4ea-1ffd-4352-b55a-e4b1c6006c67 /media/fd9cf4ea-1ffd-4352-b55a-e4b1c6006c67 ext4 defaults,nofail,user_xattr,noexec,usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0,acl 0 2

UUID=ea7fd64c-215d-4229-bfeb-fbe7ba50023b /media/ea7fd64c-215d-4229-bfeb-fbe7ba50023b ext4 defaults,nofail,user_xattr,noexec,usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0,acl 0 2

UUID=35249aed-475c-42b4-9769-51bad54e154c /media/35249aed-475c-42b4-9769-51bad54e154c ext4 defaults,nofail,user_xattr,noexec,usrjquota=aquota.user,grpjquota=aquota.group,jqfmt=vfsv0,acl 0 2

/media/ea7fd64c-215d-4229-bfeb-fbe7ba50023b:/media/fd9cf4ea-1ffd-4352-b55a-e4b1c6006c67:/media/8060a38c-45d3-4571-855d-3830f71cb846:/media/3b4c969a-9010-4883-a141-1758ddc8fa3c:/media/d7656b4c-0bac-4db8-825b-427a09d90b47:/media/d76b7e96-fced-4edc-a32a-cbb20968db87:/media/eec8ff77-1dc9-41a8-9e7b-f9ef53a5bf63 /media/80a7435e-d603-4935-828f-e23abd7999a9 fuse.mergerfs defaults,allow_other,category.create=mfs,minfreespace=4G 0 0

# <<< [openmediavault]

Alles anzeigen

Ryecoaaron,
Thanks a bunch for taking care of the backport of fuse and libfuse. I installed them, and reestablished a mergerfs volume with epmfs just like before and resumed the same copy operation. So far, so good. Will let you know if the copy operation completes without incident.
Really appreciate the help.
Best,
Phil

Hi,
I'm using mergerfs on OMV 2.2.4 for a few days and everything works well. I'd like to use the eplfs policy that is not present in the plugin interface for OMV 2.

I'm using a VM for all the tests.

I've edited /etc/default/openmediavault with this:

OMV_FSTAB_MNTOPS_MERGERFS="category.create=eplfs,category.search=newest,defaults,allow_other,moveonenospc=true,minfreespace=10G,fsname=mergerfs"

Then:

omv-mkconf fstab

and rebooted.

The new values were not applied at all and only moveonenospc=true is present (@default moveonenospc is set to false).

# file: .mergerfs
user.mergerfs.category.action="all"
user.mergerfs.category.create="epmfs"
user.mergerfs.category.search="newest"
user.mergerfs.func.access="newest"
user.mergerfs.func.chmod="all"
user.mergerfs.func.chown="all"
user.mergerfs.func.create="epmfs"
user.mergerfs.func.getattr="newest"
user.mergerfs.func.getxattr="newest"
user.mergerfs.func.link="all"
user.mergerfs.func.listxattr="newest"
user.mergerfs.func.mkdir="epmfs"
user.mergerfs.func.mknod="epmfs"
user.mergerfs.func.open="newest"
user.mergerfs.func.readlink="newest"
user.mergerfs.func.removexattr="all"
user.mergerfs.func.rename="all"
user.mergerfs.func.rmdir="all"
user.mergerfs.func.setxattr="all"
user.mergerfs.func.symlink="epmfs"
user.mergerfs.func.truncate="all"
user.mergerfs.func.unlink="all"
user.mergerfs.func.utimens="all"
user.mergerfs.minfreespace="4294967296"
user.mergerfs.moveonenospc="true"
user.mergerfs.pid="1359"
user.mergerfs.policies="all,eplfs,eplus,epmfs,erofs,ff,lfs,lus,mfs,newest,rand"
user.mergerfs.srcmounts="/media/80526d75-18b1-4897-93f4-d3efadc72afb:/media/5613597f-7839-4a90-b682-693fdcc4b3dd:/media/6dd23c37-48d5-44f3-9aed-0419615e9310"
user.mergerfs.version="2.13.1"

Then I 've noticed that the /etc/fstab reports both the values, those set as default in /etc/default/openmediavault and those set from the plugin interface, and I think that the latest are taking precedence.

/media/80526d75-18b1-4897-93f4-d3efadc72afb:/media/5613597f-7839-4a90-b682-693fdcc4b3dd:/media/6dd23c37-48d5-44f3-9aed-0419615e9310 /media/ba052b85-76d6-4d0d-bb41-d7ef2f50cb51 fuse.mergerfs category.create=eplfs,category.search=newest,defaults,allow_other,moveonenospc=true,minfreespace=10G,fsname=mergerfs,category.create=epmfs,minfreespace=4G 0 0

What am I doing wrong?

OMV Version: 2.2.4
mergerfs version: 2.13.1
fusermount version: 2.9.4
FUSE library version: 2.9.4
using FUSE kernel interface version 7.19

It works perfectly now.

Thanks for your help ryecoaaron.

Zitat von tomfowler

I can confirm that compiling fuse 2.9.3 from source worked for me. I was having constant disconnects using mddhfs or mergerfs under heavy load (crashplan), didn't matter.

Download fuse 2.9.3 <a href="https://github.com/libfuse/libfuse/releases/download/fuse_2_9_4/fuse-2.9.3.tar.gz" class="externalURL" rel="nofollow" target="_blank">here</a> from the <a href="https://github.com/libfuse/libfuse/releases" class="externalURL" rel="nofollow" target="_blank">Github page</a>

Transfer to your root user…

I've done this on a new, updated OMV installation. 3x 3TB SnapRaid and AUFS with MFS File System ... about 10GB file transfer over GBIT ethernet the server gets unavailable, just a hard reset get it alive again. SnapRAID isn''t doing any sync. The error must be witihn the union file system plugin.
any news / workarounds here?