Physical drive is definitely toast, as shown in picture #1.
Picture #2 shows that you have four drives in four logical RAID0s which is also not ideal.
If you install the drive into slot 2, and don’t configure a Logical Drive, check to see if it is visible in your TrueNAS installation. If so you can use the REPLACE option - and you may have a path to migrate drives one-at-a-time to “passthrough/unconfigured” mode, and then flash to IT mode.
Follow-up question: If you look at the Details of your controller on that page or others, does it give you any information about “write cache” or “cache behavior”? Wanting to ensure you are not only using safe write behavior but also that you’re getting maximum performance.
In the controller details i can see only this information:
I also noticed that if I look at the “Disk” section from the TrueNas webui there are:
sda (Pool HDD)
sdb (Pool HDD)
sdd (Pool HDD)
sde (SSD with boot pool)
sdf (Pool HDD)
The failed disk was originally “sdc” (Right?)
In the dashboard I see that I have an “Unassigned Disk”.
Is it possible that you are actually seeing the disk?
What happened is that the previous RAID0 logical disk presented through the IR mode on your card had an old serial number assigned to it by the RAID card. This S/N and UUID mapped it to sdc. When the physical disk was reinserted and/or reinstalled, the logical disk is now “dead” because it was RAID0 (no redundancy) - but the underlying physical disk is now getting passed through to TrueNAS, which is seeing its genuine S/N and has probably given it a new UUID - which is mapping to sdf.
Replacing the missing sdc with the new sdf should work; however, if you do this, you don’t want to “initialize” or otherwise define a new logical disk in the RAID BIOS as that will wipe it again.
As you’ve only used 1.6T or so on this filesystem I’d like to ask - how challenging would it be to back up all data here and rebuild again?
Edit: I’ve rehosted your images locally. Can you please upload them directly to the forum in any future posts for auto-resizing and easier access by all? You can drag-and-drop them right into the compose window.
I’ve read your note about the images, okay, I hadn’t found the right button, but now I know where it is
I can try, but I will definitely need a few days because as I was saying the physical hardware is located in a separate office far from my location.
At the moment can I try to replace the disk by mounting sdf? In order to have the redundancy necessary to give me a few days to reach the headquarters.
How do you recommend rebuilding the TrueNas to make optimal use of the system?
Thank you very much for the information, this is very important for us users with little experience on this system.
At the moment, using the REPLACE option to replace sdc with sdf should work, to regain your redundancy.
My recommendation would be to back up all of your data, flash the card into IT mode (or replace it with one that is already in that mode) and then rebuild your pool and restore the data.
You may be able to do it one disk at a time by the same process you have done here (hotplugging a drive, and then replaceing the failed disk with the new passthrough-mode one) but that will put data at risk every time unless you have a spare disk to use as the “pivot mode.”
At this point I will try to proceed as suggested and rebuild the Raid with the PRAID card in IT mode but i need more time…
I ask you one last question, can I also try the option of adding the disk to the existing pool from the “Storage Dashboard” without carrying out the “Replace” (which doesn’t seem to work)?
I would be hesitant to do a wipe - the additional layer of abstraction of the RAID card here is posing a challenge to TrueNAS being able to identify the disk correctly, so I would be concerned that disks have “shuffled” in a manner of speaking, and the loss of a second disk through a mis-targeted wipe would put the entire system offline.
No - this flow is designed for adding new disks as additional storage to the pool, not replacing an existing failed disk.
What does the output of lsblk show from the command-line? Using the </> button in the composition window for “Preformatted text” or placing the text between triple-backticks
I have other servers like this, and their controller are all between 60 and 70 degrees Celsius, this is the only one that reaches 80.
However, i have already seen operating temperatures like this in other machines and they have not created too many problems. The environment where the servers are located is air conditioned and the machines are cleaned of dust regularly.
Maybe I could add another ventilation, but it would be something about self built, and I don’t know how functional it would be
If the sdf disk is a new replacement for the one that failed, then I guess you can go ahead and do a replace, however if it is the same disk that failed that is represented natively now, then before you replace you should probably:
Do some basic functionality tests on it to make sure it is working and can be read and written to.
