UtahDeLorean
-
Posts
113 -
Joined
-
Last visited
Content Type
Profiles
Forums
Downloads
Store
Gallery
Bug Reports
Documentation
Landing
Posts posted by UtahDeLorean
-
-
Story time
Server froze up and when it restarted of course a parity check started. It was going really slow and then I started to get seek errors. Being a responsible unraid owner I purchased a new drive to replace the failing drive. I got the new drive in today. It replaced a vertical sata drive for a SAS drive in my disk shelve. Started the rebuild process, but it was going slow too and then I started getting seek errors on another drive that was vertical also. I relocated this 2nd drive to a horizontal slot and the seek errors went away. I know that according to Google that the orientation does not effect the drive.
Would this a situation of a drive going bad, a connection that's bad, or something else. My drives are in rosewell hotswap bays
-
I am going on another 24 hours. Is there a quicker way to fix this so that I can get my server up and running again?
It just keeps saying this
Phase 1 - find and verify superblock... couldn't verify primary superblock - not enough secondary superblocks with matching geometry !!! attempting to find secondary superblock... ..found candidate secondary superblock... unable to verify superblock, continuing... ..found candidate secondary superblock... unable to verify superblock, continuing... .found candidate secondary superblock... unable to verify superblock, continuing... ...found candidate secondary superblock... unable to verify superblock, continuing...
-
4 hours ago, JorgeB said:
There are what look like power/connection issues with this disk spamming the log:
Model Family: Western Digital Ultrastar (He10/12) Device Model: WDC WD100EDAZ-11F3RA0Try and fix that when possible, how are you running xfs_repair? If it's not from the GUI post the command used.
For future reference where did you find those details about the error with the drive?
-
25 minutes ago, JorgeB said:
There are what look like power/connection issues with this disk spamming the log:
Model Family: Western Digital Ultrastar (He10/12) Device Model: WDC WD100EDAZ-11F3RA0Try and fix that when possible, how are you running xfs_repair? If it's not from the GUI post the command used.
Okay I am going to shut down the server and check all of the connections. I was using the GUI, but I can use the terminal. Thanks for your quick reply
-
Just want to know how much longer I should wait. I've been a little over 24 hours on a 10tb drive.
I wished that I knew what caused the corruption to happen
-
I got 2 more disk shelf's that was gifted to me. I found some drives, but they are formatted to 4096. Would this process work the same to make them 512?
-
4 hours ago, JorgeB said:
That won't work, U.2 is a PCIe connection, you could directly connect for example a U.2 NVMe device, to connect STA devices you'd need to install a SATA controller using U.2 interface, not sure if they exist, they do exist in M.2 format, like this:
This is what I was afraid of when it didn't work and I couldn't find any thing int he Bios. I have extra space in my Disk Shelf so I think that I am just going to move the drives and that should allow better airflow in my case. The plan is to RAID these other drives that are 10k rpm drives. So I am trying to find a 12g external SAS card to link these new Disk Shelf's into my server
-
I have a hardware issue (1st world one I know) that I hope someone would be able to help me out with.
I have a Threadripper 2950 on Asrock X399 with 4 PCIe slots that I can use. 1st is for a Video card for Transcoding, 2nd is my 10gig card, 3rd for my External SAS that's running my Disk Shelf, and the 4th one for HBA to run all 12 drives in my rack.
I was gifted 2 Dell Disk Shelf's that I would like to add to my server and I would like to add a 12gig external SAS card (since these drives are 12gig)
I thought that I would be able to utilize the onboard U.2 card. Thinking that I was being smart I purchased a breakout cable from amazon
https://www.amazon.com/dp/B0912P8CS8?psc=1&ref=ppx_yo2ov_dt_b_product_details
For some reason U.2 there is not a lot of documentation online about the U.2 port and after playing with it and trying to get it to work I have given up and wondering if someone has a solution. I just need to be able to move 4 drives onto another type of connection to get this to work.
If anyone has any suggestions it would be greatly appreciated.
-
On 2/17/2023 at 5:08 PM, drwtrippy said:
Did you ever get this sorted out? I have one drive that is throwing this same error.
Yes... I let it work through and it took a reboot the next day, but it worked!
-
1
-
-
So I just noticed that it looks like it's still working in the background and see the progress of 13.99%
I will check on it in the morning
-
@DannyG I had a situation where I was formatting bigger drives that I got. I got the 1st drive formatted and swapped out a drive that was tossing out error's.
I was on the 2nd drive and it was running over night and when I check it the window was closed.
I put in the 3rd drive and the window crashed/closed while I was watching it... When I tried to run the format line again I get this
Does this mean that the format is still going or did I do something wrong? I have only complete 1 out of the 5 drives that I picked up. (I already did this on 4 other drives I purchased before with out any problems like this)
-
I have the same processor and I love it.
I only have one video card in my server and use the HW Transcoding on a EVGA 1660 and it was one of the best things that I did. My CPU stays pretty much Idle even when I have up to 7 streams going on (I have a 1Gbps connection and share my plex server with lots of family and friends). So I love having the cores of this CPU.
My buddy has the 1950x and he is looking to going to a newer Ryzen and I wish him luck. How ever he runs his server off of Windows 10 home and I think that it gives him more trouble then it's worth.
I would recommend tossing in a GPU and possibly changing some power setting in the Bios. I read that Unraid does have an option to run a CPU governor to reduce power too.
-
13 hours ago, JorgeB said:
I expect it will be still limited to 30, but it's mostly a guess.
I would like to see this number increase, or maybe an options to have drives set as hot spares that can be set to automatically swapped out. Right now I have 12 drives in my server and then I have a 24 disk shelf. So it would be cool if you have a drive go down or a drive that tosses out a bunch of errors that the server will just swap it out and put the other drive out... I know that this is wishful thinking though
-
2
-
-
On 3/30/2022 at 8:02 PM, chansearrington said:
@noski thanks!
I'm definitely not an expert in this area.
I decided to go the same route as everyone else and just get a DS4246 from here -> https://www.ebay.com/itm/202404952486
I assume that what you are talking about doing would work the same, but I don't know.
If yours works the same as mine, I only connected 1x cable from the Disk Shelf to my Server. Works fine. And I don't use the second disk manager module at all. Doing so, as I understand it, would make my disksehlf show up as two separate drives vs 1x combined one.
Still not sure if this is how I should be doing it, lol.So I ran all of the cables... It's for redundancy. I started with a 4243 and upgraded the modules (for rather cheap later on) to 4246 and noticed a big difference in my speeds... Because I had the cables daisy changed I was able to pull 1 module at a time and not loose connection to any of my hard drives on my disk shelf.
On 3/6/2020 at 8:11 AM, DannyG said:Parts used in this Build
- Netapp DS4243 Shelf (no controller)
- QUAD SAS Card: PM8003
-
SAS Cable QSFP to QSFP
If you have a different SAS card, I'm sure it will still work. This is the card that I was able to source.
-
There's no Power button on these shelves, they just turn on when you give them power. from my testing,
you only need 1 power Supply to be on and all 24 drives will still work. the other power supplies are only to be used for redundancy (ideally from different power sources) and is over kill in a home lab. - EDIT: I was getting "random" drives disconnecting form my array, to re-add the drives, you had to rebuild the array which took a long time and put my data at risk. In the logs it was found that my drives would go missing each time my UPS did a self check. the fix was to use 2 power supplies, I put them on their own UPS.
Here are some power stats for fun:
Empty Shelf with all 4 Power Supplies ON. (207 Watts)
- The Full Shelf, with 24 Disks on takes:
- 305 Watts with 1 PowerSupply on.
- 323 Watts with 2.
- 348 Watts with 3, an
-
378 Watts with all 4.
STEP 1) Install your SAS card in your unraid server. leave the server off for now.
STEP 2) - With the Netapp turned OFF; Physically plug the SAS Cable in the port with the "Square icon" on the top controller & the other end of the cable to your SAS Card in your unraid server.
- Turn on a Power Supply.
- Give it a good minute to fully initialize.
STEP 3) Power on your Unraid Server & login to its interface.
- If you're lucky. you're Done. The drives are already showing up in the MAIN menu, and you can start assigning your Disks.
if you don't see your Drives... Keep reading.
STEP 4) Go to TOOLS > System devices.
- Do you see your disks? (yes = good) (No? Go to Step 5)
- Your Disks don't have a "Size"? (Let's fix this)
STEP 5) Sanity Check - Check the Logs
- Go to: TOOLS > SYSTEM LOGS
- CTRL-F (find), search for "Netapp"
- Finding "Netapp" confirms that your card is working. (Great!)
- Some Netapp drives are formatted with 520b sectors instead of the needed 512b. we need to reformat them for Unraid to see them properly.
STEP 6)
- Click the Terminal icon
- Run this command to download the formatting utility:
wget http://slackware.cs.utah.edu/pub/slackware/slackware64-14.1/slackware64/l/sg3_utils-1.36-x86_64-1.txz- Run this after the download to install sg3_utils
upgradepkg --install-new sg3_utils-1.36-x86_64-1.txz- Use this command to show SAS HDD's
sg_scan -ithis command to format. PLEASE READ /dev/XXX should be the HDD in the NetApp that needs to be formatted. (AKA double check before you format the wrong drive)
sg_format --format --size=512 -v /dev/XXXFrom my testing, I was able to open 24 terminal windows and format all 24 drives at the same time. This might have slowed things down a bit.. not sure.. but I just went to bed.
STEP 7)The following morning, I woke up, went to MAIN, but still didn't see my disks
I REBOOTED the System.... and the 24 Disks appeared.
Checked the Logs... no more errors... and I confirmed it was formatted in 512 sectors now. yes!
I just picked up three 4Tb drives and had to reformat them. I am on the most recent unraid build and when I tried to run the wget I got an error about there not being the file. So that might need to be updated.. I ended up figuring out which drive it was and ran the sg_format command. I have 2 real questions. Once ran do I need to keep the terminal open until it's done or is it a situation that can just be closed and it will finish when it finishes?
Also is there a way to speed up the process? Looking to get more and bigger drives as time keeps moving along
-
2 hours ago, ljm42 said:
What url is My Servers trying to use for Local Access?
I'm guessing there is an issue with your "Local TLD" setting. Please upload your diagnostics (from Tools -> Diagnostics)
Something changed when I moved to my new router...I wished that they made a good 10gig router that I didn't have a get another mortgage on my house for lol
-
1 hour ago, ljm42 said:
If you don't want to use https to access the webgui, go to Settings -> Management Access and set "Use SSL/TLS" to No.
Yea I never changed that... I only changed Routers and set it as static
On a side note... I can't access my server through the "My Servers" tab... It shows as Local access, but it doesn't pull up
-
I don't know why but I had to put https before my server name/number
-
Okay I don't know what's going on... It shows up as the correct IP... I could access it, but now it's gone back to unable to access (just the gui)
-
Never mind... I deleted it and set it back up and now it's working (I swear that I did that once before already)
-
So I had an Edgrouter lite, how ever it seemed to give up the ghost... I purchased a Mikrotik CRS305-1G-4S+IN thinking that It would be able to handle my 1Gbps connection. How ever I guess that it only handles 10gig as a switch (So I am going to run it as a switch now) and I picked up a Mikrotik RB4011iGS+RM (I was hoping to go all 10gig, but 10gig wan is still really really $$$$)
Still the plan is to remove my gate way can connect the router to my Fiber connection.
TL:TR My new router when I try to put my unraid server as a Static IP address all of my dockers can still be accessed, but I can't access the Unraid GUI now. I had no problems on my CRS305, but the RB4100 is giving me an issue when I set as static on the router
-
25 minutes ago, jonathanm said:
Individual files can't span disks. I suspect you are using most free, which ends up scattering various thumbnail and nfo files all over. Consider this scenario. All disks are even to within 10's of MBs, so the main movie file gets written to one disk, then it no longer has the most free so the ancillary files that are in the same folder get written to various other disks.
Most free allocation method means the drives are constantly filled round robin to stay even, effectively spreading the files all over and ensuring that any browsing will spin up all the drives. Good for free space OCD, very bad for performance and folder organizational OCD.
For my OCD what would you suggest then? Honestly I was swapping out drives for 10tb drives... Then I got a Disk Shelf and increased my drives to the max. My system has filled the 10tb drives to 88% capacity which drives me nuts cause I have other drives that are still sitting at 1%
-
1 hour ago, trurl said:
Not entirely sure what you mean, but I suspect you are misinterpreting something.
When you look at what's on the different Disk's it will have just like 1kb on multiple disks
But the actually file for the movie will really only be on 1 disk
-
11 hours ago, tjb_altf4 said:
Aesthetically it bothers me, but functionally it does what its meant to.
I find that many of the file are on just one drive, but there is a reference file on the other drives
-
Does it bother anyone else when there is movies being placed on multiple drives like this
Seek Errors
in General Support
Posted
Story time
Server froze up and when it restarted of course a parity check started. It was going really slow and then I started to get seek errors. Being a responsible unraid owner I purchased a new drive to replace the failing drive. I got the new drive in today. It replaced a vertical sata drive for a SAS drive in my disk shelve. Started the rebuild process, but it was going slow too and then I started getting seek errors on another drive that was vertical also. I relocated this 2nd drive to a horizontal slot and the seek errors went away. I know that according to Google that the orientation does not effect the drive.
Would this a situation of a drive going bad, a connection that's bad, or something else. This drive are in rosewell hot-swap bays, but I have other drives in external Disk shelves