John_M

I wrote the instructions on the basis that they use only functions already built into unRAID. They are easy for a beginner to follow and nothing needs to be downloaded or installed. I've tested the procedure and it does work but it isn't necessarily the fastest method. There are several ways to achieve a goal and they all have pros and cons. The OP wanted to physically replace his cache disk. It's safer to power down and I'm still not convinced that unRAID completely supports hot-swapping. Some people have reported success, others failure. So my general advice would be to power down. If advanced users find they don't need this step then so much the better but there's nothing to be gained by confusing beginners.

Since your power supply has a single 12 volt rail there's no balancing to be done. All the yellow 12 volt wires meet at the same point inside the supply. The only (minor) consideration is to make sure you don't draw more current down that single 12 volt wire than it's rated for. With what you're proposing to do it won't happen but you don't want to be connecting splitters to splitters to splitters, for example.

Using a network share (even Apple's own Time Capsule) as a Time Machine destination is never as reliable as directly connecting the disk to the Mac. The difference is that in order to implement Time Machine across a network a sparse bundle disk image has to be created on the destination, whereas a directly connected hard disk can be formatted appropriately. Sparse bundle disk images consist of many many small files and they have proven to be rather fragile in practice. What I do is directly connect an external USB-connected disk to each of my Macs and let Time Machine write to it directly. It's simple and reliable.

A few thoughts: Ignoring unRAID for a moment, how much RAM does the BIOS see? Have your tried removing one pair of DIMMs, then the other pair? Is the RAM you're using on the qualified list for that motherboard? Are you trying to overclock? If it's an old gaming computer, you may have some BIOS settings that are undesirable for a server. Have you run memtest?

What SAS/SATA card are you using? Do you see the correct amount of memory if you remove it? There's a strange problem that can cause half your memory to be ignored if you have all four slots filled. The remedy is to mask off two pins on the PCIe connector. See here: https://lime-technology.com/forum/index.php?topic=12767.msg159603#msg159603

The only logical answer is that you changed it yourself, accidentally, while you were altering permissions. It's very easy to click on the wrong item if your attention lapses for a moment and it's my biggest gripe about using a GUI for administrative tasks.

Is it surprising ? Perhaps you could explain why instead of giving an unhelpful reply. I'm sorry you found my reply unhelpful. I can't fix your problem but I thought I could give you the pointer you were asking for to fix it yourself. I can't give you an explanation because I didn't write the plugin but I assume it polls the mounted devices when the page is refreshed. Try unmounting your devices before switching them off.

Do you unmount those devices before you switch them off? If not then it's not surprising that UD hangs for a while.

It would, because /var/log is a 128 MiB tempfs which exists in RAM. Therefore it was cleared when you rebooted. The problem was that it had filled up and there was no space left. The question is, why was it being filled up? You'll have to wait a while and see if it starts to fill up again in order to investigate. You can monitor its usage on the GUI Dashboard page under System Status, flash/log/docker: the middle gauge. EDIT: I've just seen your other thread here. The error messages you mention there are filling up your log, though it may not be the only cause.

The file you want is /boot/config/plugins/unassigned.devices/samba_mount.cfg It's easily editable.

1. gfjardim's script is included with the plugin. If you have the latest version of the plugin you have the latest version of the script. If you want to use any other script you need to add it manually. 2. The fast post-read option only applies to bjp999's script. 3. You're trying to hot-plug disks, which isn't officially supported, but see here: https://lime-technology.com/forum/index.php?topic=53473.0 4. I left a parity build (P and Q) and two pre-clears running overnight. The parity completed without problems and the pre-clears are progressing well.

You can make the Unassigned Devices plugin automatically run a script whenever you plug in an external USB disk. Several people use that as an effective backup method. If you want to put the disks inside your server so that they are permanently mounted using UD you could use the User Scripts plugin to schedule regular backup tasks. See http://lime-technology.com/forum/index.php?topic=49992.0 and http://lime-technology.com/forum/index.php?topic=50416.0

You're welcome. Yes, it's best to rebuild the docker image before it becomes a problem. It's a simple task as all your settings are remembered.

I updated and it's looking good. Thank you.

The "Use cache disk: Prefer" option is only available in unRAID 6.2 and later. It makes the mover work in reverse, moving files from the array to the cache.

There are some instructions for old versions of unRAID here: https://lime-technology.com/wiki/index.php/Replace_A_Cache_Drive But since you're running a very new version, how about this: 1. Stop your dockers and VMs. 2. Disable docker and VMs in Settings, and switch off array auto-start. 3. Set the shares that are present on your cache disk to "Use cache disk: Yes". 4. Run the mover. 5. Check that the cache disk is now empty. 6. Stop the array and unassign the cache disk, then power down. 7. Replace the cache disk. 8. Power up and assign the new cache disk, then start the array. 9. Format the new cache disk. 10. Change the shares that you want on your cache disk to "Use cache disk: Prefer". 11. Run the mover. 12. Check that the cache disk is no longer empty. 13. Enable docker and VMs in Settings, and switch on array auto-start if preferred. 14. Start dockers and VMs.

Thanks for this. I'm interested but cautious. I need to do some reading before I install it.

I'd like to give dgaschk's original idea a bump, since modifying the /etc/rsyncd.conf and /etc/inetd.conf files on boot-up has traditionally been done via the go file, a practice that is now deprecated.

A teeny tiny buglet in Dynamix System Stats which is hardly worth mentioning but ought to be easy to fix: the "Normal usage" legend for the green bars used in the Disk Stats tab is displaying with reference to the configured critical disk usage threshold, not the configured warning disk usage threshold. In my case I have warning set to 80% and critical set to 99% so it reads "Below 99%" when it ought to read "Below 80%". A screen grab explains it better:

If you exported the checksums you can search the resulting hash file for your missing photos, since it consists of a list of filenames with associated checksums. If you didn't export the checksums the copy held in the extended attributes will have been lost when the files were deleted.

That's fixed now. Thank you.

The view from Unassigned Devices looks like this. As I can't clear the red message the Mount button is greyed out. The eye icon works but the red cross doesn't. EDIT: It seems to be a browser (Safari) issue. I cleared the cache and history but it still didn't work. I tried Chrome and it works.

Testing the 2016.10.26 version, I tried the gfjardim script to "Verify MBR only" a disk that has an XFS filesystem on it. As expected it failed the unRAID signature test. But how do I clear the red failed report? Clicking the red cross does nothing.

Thanks. That fixed the installation problem, but there's a dependency that's unresolved: Same for host and nslookup .

Found your issue. Will be fixed in the next update. Excellent news! Thank you.