xrqp

20 hr later update. Nzbget download rate slowed to a crawl, even though my internet is tested to be fast. I tried again with the removal of the nzbget container, then re-installed the container using unraid "previous apps". And now my nzbget download is fast. I am guessing it will last for a 12 to 24 hours, then I may need to do it again. At the same time i updated unraid from 6.9.1 to 6.9.2, and all my containers needed updates too, so did that too. Maybe that will help keep the download rate high.

Hey. I did remove the nzbget container, then re-installed the container using unraid "previous apps". And IT WORKED GOOD! I did not loose my download queue. And it is downloading at full speed. What a relief! Unraid and my dockers frequently give me a good scare, but usually I can fix it, with time and effort. Is it fair to assume it will become more bulletproof as the years proceed? Sometimes I wonder if running emby, with tv antenna, and all the *arrs, and doing disk rebuilds, and parity checks, and doing preclears - if it overloads it and causes problems?

Hey RecycledBits, thanks for your post. I had not thought about the energy saving of one 18TB vs two of 9TB, but it makes sense. I would give more weight to your experience since you had a lot more drives than me. Were many of your 8TB SMR used in unraid array? Not sure if that makes any difference. Other posts said SMR worked very well in unraid.

My nzbget is not downloading after i did a successful parity check and disk rebuild. I don't know if that was a coincidence. It seems to otherwise be working, like I can add nzb's to the queue. It says "downloading" but the speed is zero. I checked my internet was good, and did power cycles of unraid, but same problem. I tried nzbget "restore" to a backup that is a few months old, but same problem. Now I want to try removing the container, then re-install the container using unraid "previous apps". If I do that do I lose my 360 item queue that took days to build? Also I am open to any other suggestions. Thanks.

Maybe what you say is because if we zeroed it in the array, the parity would have to be writing the changes as well as the disk being zeroed for removal. Is that what you mean? And is that about half speed of the normal method of writing to just one disk (writing the emulated disk on to the new disk via parity sync)?

Sounds good. Thanks. I am still on 6.9.1. I will soon upgrade to 6.9.2, after the parity rebuild 😄. The 18TB parity rebuild looks like it will be about 44 hours total, only 2 hours left to finish.

It takes 2 or 3 full days to parity sync an 18TB drive. Can this be a new feature request, that we can stop array and continue parity rebuilds? Preclear allows reboots and can continue where it left off. I have frequent problems with things not working right and needing to reboot. For the last 2 weeks I often don't make it to 2 days before I need a reboot. I don't know if it is because I run sonarr and radarr and nzbget with hundreds, or thousands, of downloads queued or grabbed, or because I used the cheap shucked 8TB drives that keep having problems, or if unraid is not very stable. I don't know, so I make guesses.

After me move all the data off a drive that we want to remove, we need a way to zero that drive in the array, so we can remove it, and not need parity rebuild. So we have continuous parity protection. Or is there a flaw in this idea?

I had checked each disk to include in user shares. Maybe I forgot I did that when I added disks before. I will change it to be no disks checked, so all are included automatically. Regarding the "may" vs "will", for me it made the disk into a disk share, without me doing any checking of any check boxes. But as time goes on, I may understand it better. Thanks very much for the extensive help in your replies.

I have had disk errors and disk failures lately, and it seems to be with 8TB SMR drives (mostly shucked). On-line I only see info about how slow they can be. Can't find much about failures. But I think they fail more too, and fail within about 2 years. I am regretting them.

"Settings->Global Share Settings", "Enable disk shares:" was set to "auto". Now is set to "No". Then I could pick disk 7. So I am back to normal. Why did I not have this problem when I added disk 4, 5, and 6? Help says "If set to Auto, only disk shares not participating in User Shares may be exported." Question 1) If I add a new disk, does it immediately go to disk share before I have a chance to put it in User Shares? Question 2) Help says "may be exported", but for me it was exported. Why does help use the word "may".

I found this in the unraid docs: I guess I need to wait for my mover to finish, then stop the array, then see if I can fix it there, at "Settings->Global Share Settings". I did not do anything there previously to cause this problem. The only thing unusual, is Drive 7 used to be my parity, before I got a new parity.

I added drive 7 recently. It looks good on the "main" tab. But when I go to a share, I cannot pick the drive. Here is the 2 screenshots, and diagnostic zip: Notice below, I can only pick drives 1-6 for a share. Why dos it not show Disk 7? tower-diagnostics-20210908-2258.zip I just realized it shows as a Disk share. see below. I don't know how it happened. I am guessing I need to stop it from being a disk share. How do I do that?

You are correct, this Exos is CMR. Here it the Amazon listing title: Seagate Exos X18 18TB Enterprise HDD - CMR 3.5 Inch Hyperscale SATA 6Gb/s, 7200 RPM, 512e and 4Kn FastFormat, Low Latency with Enhanced Caching (ST18000NM000J).

Yes. I did not even notice until you mentioned it, but my order record shows it is from HyperHawk. I would be upset about the switch (if it was a switch or hidden), except I got the 5yr factory warranty, so it is OK. Ordered Aug. 30, recvd Sept 6, warranty to Sept. 30, 2026. I think they go by mfg date plus 1 month to sell, plus 5 yr., or something similar. It is difficult to know if you are getting factory warranty, or dealer warranty which is more risky. In this case, if Seagate factory did not show for the serial number I received, I would have returned it. But it is scary since the listing does not say "factory" warranty. Is Exos an SMR? If it is not SMR, then that's one more reason it is a sweet spot for me.

What hard drive purchase makes sense today, Sept. 2021? For me, it is 18TB Seagate Exos, w 5yr warranty. Amazon $400. ($22/tb) Shucked drives? Currently I am seeing no bargains. Back in July Aug. 2021 Costco had Seagate 8TB for for $130 ($16/tb). Now everywhere seems to be $170 ($21/tb). No warranty when shucked. Internal w 2 to 3 yr warranty: Also currently 8TB for $170 ($21/tb). Why do I prefer to pay $22/tb for an 18TB vs waiting for an 8TB sale ($16/tb)? Because: 1) 5yr warranty 2) Lot more TB in a single unraid enclosure (tower or case). Of course all of this may not apply to those who do not need more storage to fit in their enclosure. I only buy Amazon as last resort, since power corrupts. If anyone knows what is going on with the Amazon seller "Seagate Store": Is it the same as "Seagate Technology Inc."? When I did online warranty check on the serial of the 18TB I received, it did show 5yr factory warranty. But I am curious if this is a way for the factory to circumvent the dealers and compete with their own dealers? My guess is they use a loophole in the dealer agreement, by creating "Seagate Store" as a company, and we do not know if it is the same owners? I gave up on trying to figure out which of the 3 major makers make more reliable drives. So I buy the cheapest which tends to be Seagate more often.

I went ahead, and put the 6.9.1 backup on my new "fast" flash, 32GB Samsung Fit Plus. My system does not seem any faster, for shutdown or reboot of unraid. Roughly the same. My prior one was Sandisk 16GB UltrFit. It was USB3 also.

I want to boot faster. Seems like I need to boot or re-boot often. Does not faster flash speed make faster boots? "USB3 drives are far more prone to failure than USB2 ones" Is it still that bad after all these years? I was going by SpaceInvaderOne video about the flash he tested. He only used a standard flash test app, not unraid test. He made a guess that the pro-USB2 (anti-USB3) for unraid may have been long ago, and may be is no longer true. Don't forget my other issue: I cannot do a flash backup since I went from .1 to .2

on the unraid tower, the flash is 16GB. and Ram is also 16GB.

I am trying to upgrade my unraid flash hardware to a faster one. I am using Win10 PC, and unraid web interface, and go to Main, then Flash, then click "Flash Backup" button. It goes to my Win10 downloads folder, but it stops and ends with the file name "Unconfirmed 551776.crdownload". I rebooted unraid and tried again w same result. Then rebooted Wind10 and same result. I recently updated to Unraid 6.9.2, but before I updated from 6.9.1 I was able to create flash backup. It created as file name "tower-flash-backup-20210824-1742.zip". What should i do - go back to my last backup of 6.9.1 ?

trurl, Thanks. What you suggested sounds like something I can do.

I tried to use "Clear an unRAID array data drive" but I couild not figure out how to put files on the flash drive. I clicked "Add a new script" so I wa able to give it a name, and I think it automatically created a folder on the flash drive in the config/plugins/user.scripts/scripts folder. I downloaded the "description" and "script" files to my Windows 10 machine. But I don't know how to put them in the unraid flash drive in the config/plugins/user.scripts/scripts folder. When I use Krusader, and go to FLASH it says, "Error: Cannot open the folder /FLASH". And even if i could open it, how do i move the file from Win10 to unraid flash?

Both repairs were on the same disk #5.

I was able to do successful xfs repair again and I got my music back. It was a hard drive problem. I do wonder if Unraid is not able to deal with hard drive issues that other software can deal with, since I have much fewer hard drive problems on Windows. Just one anecdote.

Yes. I had a lost+found folder from doing xfs repair about 8 months ago. It seemed too laborious to go thru. Also, since it has so many files and folders, it took time just to open the folder and subfolders, in windows Explorer. I assume the xfs repair yesterday may have added more to the old lost+found folder. I guess when it creates files there, they keep their old creation date, so I could not separate the ones from the 2 repairs. My guess is that it is not worth it to save them, but I need to search for info and read more about it. I don't know Linux, but I may try the file command. Thanks.