-
Posts
16136 -
Joined
-
Last visited
-
Days Won
65
Content Type
Profiles
Forums
Downloads
Store
Gallery
Bug Reports
Documentation
Landing
Posts posted by JonathanM
-
-
2 hours ago, csimpson said:
However, if I add another 4TB drive, (not parity) would I then have 12TB of protected data?
The math doesn't add up for me.Hoopster summed it up quite well, but I wanted to stick my .02 into the discussion to hopefully clear this up a little more.
Parity doesn't hold any data. Period. It's not a backup. Period.
It contains the missing bit in the equation formed by adding up the bits in an address row. Pick any arbitrary data offset, say drive1 has a 0, drive2 has a 1, drive3 has a 1, drive4 has a 1, so parity would need to be a 1 to make the column add up to 0. Remove any SINGLE drive, and do the math to make the equation 0 again, and you know what bit belongs in that column of the missing drive.
So, you can protect ANY number of drives, and as long as you only lose 1 drive, the rest of the drives PLUS PARITY can recreate that ONE missing drive. Lose 2 drives, and you lose the content of both, but since Unraid doesn't stripe across drives, you only lose the failed drives.
Unraid has the capability to use two parity drives, so you can recover from 2 simultaneous failures. However, the second parity is a much more complex math equation that takes into account which position the drives are in, so it's a little more computationally intensive. The extra math is trivial for most all modern processors.
-
51 minutes ago, Napoleon said:
Any reason why no one mentions this PSU?
New, unproven, expensive?
Advertising looks great, do you have any links to third party real tests?
-
21 hours ago, tejasgadhia said:
it doesn't appear any new files are being written to the drive, so I assume Unraid knows that the drive is not healthy.
That's not a thing. Unraid will quite happily continue to use a disk slot even if the drive fails a write and is disabled.
-
1 minute ago, Inch said:
I've tried plain bridge mode and custom network with the same results. I've just switched back to bridge, readded the sonarr port (specified 8989 in both) and it still doesn't work yet Radarr does still.
Strange. I'm out of things to try at this point. Maybe someone else will have some ideas.
-
2 minutes ago, Inch said:
Tried that, still doesn't show it.
Probably because you are using a custom network for delugevpn instead of the default bridge. Binhex doesn't support anything but plain bridge. Doesn't mean you can't make it work, but it can be challenging.
Maybe the radarr port was added while you were in plain bridge mode?
-
6 minutes ago, BreakfastPurrito said:
By recreate, do you mean plopping in a docker.img from a backup or is there a specific procedure?
Specific procedure, pretty much just what I said.
Delete the image file, start the docker service to create a blank one, add your custom networks, go to previous apps and tick off all the ones you want to put back.
-
-
1 minute ago, Inch said:
Advanced view for the VPN image.
click edit on the 8989 entry and post just that part.
-
There may be other ways, but the quickest way I know is delete and recreate the docker.img file.
Be sure to create your desired networks BEFORE going to previous apps and selecting everything you want to reinstall. It should only take a few minutes, and everything will be back the way it was.
-
5 minutes ago, blackbird2150 said:
I can't tell if that is my problem.
Nope. Forward vs. reverse = no drives detected with wrong cable.
-
25 minutes ago, Inch said:
Added another port for `TCP` under the same deluge vpn docker for port `8989`
Show a screenshot of that in advanced view.
-
18 minutes ago, Inch said:
Under the same Sonarr docker image, I've set the network type to `none`
If you set the network to bridge without changing anything else, does it work as expected, other than not running through the vpn?
-
17 minutes ago, DeDude said:
I dont know why Unraid is using Anti-Cheat/DRM techniques to identify a HDD - just do it with the volume ID maybe?
Couple factors. Unraid runs in RAM, which effectively means each boot is like a new install. The way linux handles drive identification normally is /dev/sdX or a variant, where each drive that is detected gets the next designation. There are so many things that can change that designation from one boot to the next, and also needing the ability to successfully boot on widely different hardware, the choice was made to identify by something that is supposed to be unique and not modifiable by something written to the drive.
Honestly, this is the first time I remember seeing a spinning rust hard drive not give a unique serial number when attached directly to a compatible SATA controller. Many times USB bridges or controllers can manipulate what is passed to the OS, so that's fairly common to need a plain vanilla SATA connection.
-
1 hour ago, Gico said:
Zeroing is on it's 5th attempt after 4 failed ones.
Does it always stop after 20TB? If so, maybe try a different controller? Maybe the one you are currently attached to has a problem with 20+TB.
- 1
-
2 minutes ago, jul9000 said:
how do I get rid of these tho?
Stop and start the array, or shutdown and reboot. Can't remember if just stopping and starting clears the error column, but I think it does.
- 1
-
1 hour ago, hardcore_2031 said:
I am not able to upgrade from the GUI. What am I supposed to do?
Upgrade using the manual method, download the 6.12.10 files, extract and overwrite all but the config folder.
Make a full backup of the USB before doing that, so if something goes sideways you have a copy of the config folder.
-
14 minutes ago, jhughes2466 said:
This is still happening.
On 1/15/2024 at 9:16 PM, novirium said:3 times now I've had new drives seemingly fail to format when adding them to the array.
Are you the same person? If not, please start your own thread and attach your diagnostics. If this is still you, attach new diagnostics covering the time period when this happens.
17 minutes ago, jhughes2466 said:I can clear them / kill the partition, format, and rebuild..
Rebuild negates all previous actions you listed, the emulated disk includes the format.
-
Memtest is only definitive if it finds errors, sometimes memory produces errors only under certain loads.
Try running with only 1 stick and see if the disk errors change, then switch sticks and repeat.
Also, make sure you disable any XMP profiles. XMP is overclocking, regardless of whether the motherboard and memory says it should support it.
-
How many hours of memtest have been run in the current config?
-
24 minutes ago, Milvus said:
what exactly does it mean?
It means recovery from corruption can be impossible with encryption in the way.
Corruption can happen with hardware errors, like bad RAM, cables, or power issues. The problem is, you don't know it's going to happen until it does, and RAID (of any sort, not just Unraid) can't always compensate, meaning unless you have complete backups, you will lose data.
Unraid or any RAID can't help with file deletion or overwriting good data with bad, so backups are always needed, but with encryption, the recovery options are even more limited, so backups are even more necessary.
If the data is important enough to encrypt, it's important enough to keep multiple copies in multiple locations.
-
UrBackup is what I use. Discussion here.
-
8 minutes ago, Jebberino84 said:
I did mention in my first post that I already tried safe mode
There are many different options in Windows safe mode. You didn't mention how many you tried, it implied just the single option (with internet)
-
1 hour ago, professeurx said:
I'm speak french sorry ,
-
2 hours ago, sannitig said:
I thought the wiping of the disk was the same as formatting.
Nope. Wiping the disk writes all zeroes, removing any traces of files or filesystem formats. Think of a format as a filing cabinet with drawers and folders. It allows files to be stored in an organized fashion so they can be easily retrieved, as opposed to just tossing the files on the floor in an empty room. Filesystems take up space even when there aren't any files stored.
- 1
Moving from an older Windows 10 system to Unraid. I'm terrified.
in Lounge
Posted
It's more correct to think of the USB stick as firmware with space for storing changed settings. Unraid loads into and runs from RAM, it only touches the USB stick when you change settings.
Container appdata and executables should live on a SSD or multiples for redundancy, separate from the main storage Unraid array. Legacy documentation and videos will refer to that storage space as "cache", now it's more properly referred to as a "pool" of which you can create as many as make sense for the desired speed and redundancy.