JonathanM

I could be wrong, but I think since Unraid changes the md driver you have to use a VM. Other than that, I agree 100%, make a raw image, and play with that rather than the real drive.

Be very careful with the amount of stress you put on the circuit board. Ideally you should be pushing directly opposite the CPU with a spring force of some flavor, so the circuit board is floating with respect to the hold down screws. The pictures you posted appear to show the only force applied to the CPU - heatsink interface is from the tension created screwing down the board. Perhaps a pair of 3D printed blocks with a spring inside? That way the CPU - heatsink are clamped by the case screws. Ignore me if I've misinterpreted the pictures and description.

https://forums.unraid.net/topic/74460-external-sas-storage-101-query-not-guide/?do=findComment&comment=686248

Sadly due to the apathy of what seems to be a majority of people, it will forever remain unsolved.

🤣🤣 What is this thing you refer to? Release Notes? Who reads those? 🤣🤣 That's sufficient, more would be redundant. Is the quoted text also in the inline help system next to the exclusive setting? (I know, I know, spin up a trial and look for myself)

So content manually added to the sharename folders on a foreign pool will be hidden until the array is restarted? That's fine, I just wanted to confirm the behaviour so we can help people who will inevitably have a container writing to the wrong location.

When is the additional check evaluated? Only when an attempt is made to change the status from NO to YES? What happens if the share config is manually manipulated? What happens if an exclusive access share has content added to another pool? Does the check happen every time an exclusive : YES share is mounted? If so, when content is added to the share folder on a "foreign" pool, is the content then made available when the array is stopped and restarted forcing exclusive to NO?

Quick clarification please, for the minority of us that have leveraged "only" and "no" in unconventional ways, i.e. to access all the content of both the array and pool but NOT automatically move things around, manually moving things as needed, does the "additional check" force exclusive access to NO, and continue to work as before showing a fuse mount of array and pool content? I just want to be sure before I get a nasty surprise where most of my VM's suddenly lose their images. I currently have domains set to cache:only, so new VM's get created on the pool, but if I don't plan to use the VM very often I'll manually move the vdisk to the array, and move it back if I need to.

https://forums.unraid.net/topic/92285-new-use-cache-option/?do=findComment&comment=854925 https://forums.unraid.net/topic/94080-individual-mover-settings-for-different-cache-pools/?do=findComment&comment=869566 I've been advocating for this change for a few years now.

When the system detects the drive has been spun up, it triggers a SMART read. So you need to figure out what's accessing the disks, the SMART read is a result of the spinup, not the cause. Try safe mode to rule out plugins, disable dockers and VM's, audit network devices to rule out periodic share accesses. Obviously the nuclear approach of turning everything off and reenabling one set of things at a time until the problem occurs is the best method, but that might not be desirable, so you may have to just work on bits at a time. The open files plugin may be helpful, but sometimes the accesses are so brief it's hard to catch that way.

You missed the point. Yes, you need extra SATA connections. You need to get a card or cards that has enough connections without a port multiplier.

Water coolers are great for PC's that are only on when someone is nearby and using it. I would never trust one to keep working 24/7/365 unattended. The consequences of unattended failure could be very bad.

I bend the tab at the top straight, then screw it down to one of the slot cover screws. It sticks out over the motherboard, typically at the very bottom or side of the case. Very secure.

I prefer the cables that are meant to be put in a rear slot, I screw it inside the case instead, bending the bracket to fit if needed.

Perhaps it would be better to build a lightweight PC based router with your tunnel and failsafes built in so your whole network is available with the proper firewall rules.

It is sometimes possible with valid parity to emulate and rebuild one drive at a time if the only issue is the partition offset. If the drives are up and reading properly in the RAID machine, and parity is valid, you could try to move ALL the array drives to the new machine, keeping the same assignments with a new config, then after Unraid shows all green balls you could pull one drive and see if the emulated drive mounts properly, if it does, rebuild to a physical drive. However, this assumes you are using all the same drives, which doesn't seem like the case here, so your safest option is copying over the network from one server to the other.

With older USB drives it wasn't as much of a factor. Modern high density USB drives seem to heat up much more. The higher transfer rates seem to generate heat quickly. USB drives aren't really designed with our use case in mind, so we have to be super selective to find a good match that endures nearly as well as the older drives.

Super small = heat buildup = dead drives Go with a large form factor with lots of metal shell, and put it inside the case where it can't get disturbed. If your motherboard doesn't have a USB port internal, motherboard header to USB socket adapters are cheap and easy to fit.

https://github.com/binhex/documentation/blob/master/docker/faq/vpn.md See Q4. Binhex's VPN containers are locked down, you must add exceptions for each IP that will be allowed to access the webui. Either list each network you are going to be accessing it from as above, or set up a reverse proxy where the container sees the accesses as local.

I believe you would need to add the remote WAN addresses you wish to access from into the LAN network container entry.

Definitely, those cards are designed for rack servers, where airflow is guaranteed over all the slots. If the card is used in a desktop, you must add forced airflow over the heatsink somehow.

I have my backup server at my business, and it's on 24/7/365 running urbackup.

Verify the disk type defined in the xml is correct, I seem to remember someone saying it reverted to raw instead of qcow or something like that.

I've seen your other posts, and from what I can tell, it works fine if you use another name for the container. Why is that not an acceptable work-around? I get that it should work without doing that, but if you have a working solution, I doubt anyone will be super excited to spend a bunch of time figuring out what's stuck in your specific system.

Safari is known to be randomly problematic for some subset of folks. Can you recreate the issues in a chrome based browser? Firefox almost works, but has a nasty habit of popping up a "resubmit information" dialog, which if you cancel, the browser works great, if you accept, it causes all sorts of issues.