How do I know if a server drive is dead through TrueNAS?

I am new to using TrueNAS and have it setup as a shared storage for an Avid Media Composer editing bay. My server is located on a different floor of our building that is not easily accessible at all times of the day. I am trying to setup some sort of alert system so I am informed if one of the drives in my server is going bad or dies completely. Is there a way to do this through TrueNAS?

You can set up e-mail notifications for various errors, like if a pool goes from healthy to unhealthy, or if a disks failed a smart test.

1 Like

Setup email alerts in the GUI. You didn’t tag your version I link the documentation for scale, switch to core if you use core.

And this is great to get information about your smart test if scheduled regularly

Thanks for your replies. I should mention that the server is not connected to the Internet, nor will it be. Is there at least a way to tell if a drive has gone down through the TrueNAS GUI? I was told by the technician who setup the server that it doesn’t do this and we’d need to get the iLO setup, but I’m wondering if this is true?

Even if the GUI just told me it had detected a drive issue, that would at least prompt me to do a physical check of the server hardware and I could see the lights on the drives to see if any of them are bad.

The version I am using is: TrueNAS Core - 13.0-U6.1

Possibly it’s also relevant to mention the whole server is setup as a RAID 60, so TrueNAS appears to only recognize the storage as 1 drive.

1 Like

If you have access to the truenas gui and there’s a Problem with a drive there’s going to be a notification in the top right corner. You can click on that message and it will give you either a rather cryptic error message or a straight forward one like "there are read or write errors on diskx

There is a bell icon in the GUI where it display any alarms.

It would be better to look into a local mail service or something similar, maybe write a script to get notificationd to some other machine you check daily

Oh no…

Do not use hardware raid with ZFS. Post your complete hardware for recommendations.
It would be best if you could create a raidz pool and restore from backup.

1 Like

Welcome to the forums.

I’ll reinforce @chuck32 warning to you. This is really critical - TrueNAS is intended to take control of the discs with ZFS managing the redundancy. Your hardware spec is most important in order to determine if your current disk controller has the right functionality to be used with TrueNAS.

Very bad design, and that will prevent TrueNAS from knowing anything about the state of any of the individual disks.

Okay, okay. I definitely appreciate all the feedback. You’ll have to bear with me a little here because I am the end user (Editor) and very new to ZFS and just learning the terminology now…

So as I understand it, using a RAIDz, TrueNAS will see each of the virtual devices and will alert me if any of those has an error–which would then lead me to checking the hardware itself and replace a drive in the server? If that is the case then that certainly makes sense setting it up that way and I’m not sure why my system technician did it with a hardware RAID (I will have to go over this with him).

Are there any other issues I may run into if it’s kept as a hardware RAID? Or is the main benefit that TrueNAS can then properly monitor the drives? Obviously now that we’ve been using the server and have about 30 Tib of media on it, switching over will be a bit of a task…

To my knowledge, my complete hardware is as follows:
TrueNAS Core 13.0-U6.1
HP ProLiant DL380 Gen9 G9 12LFF 2x 10CORE E5-2640V4 2.4GHz, 128GB RAM, 2x Silicon Power 128BG SSD boot drives, 12x HGST Ultrastar HE10 10TB 7200 RPM drives configured in RAID 60.

I could be wrong & smarter minds than mine will hopefully correct me; as I understand it without direct access to each drive, ZFS will not be able to fully validate data & you run the risk of silent corruption when mixing hardware raid & ZFS. I’m sure that this is a gross oversimplification.

1 Like

@dan provided a useful resource, I didn’t read it completely since I stay away from hardware raid.

But I’ve seen enough posts where the pool eventually failed and recovery was a disaster of didn’t work at all because of the interference of the hardware raid.

My personal take is using a hardware raid solution will eventually lead complete data loss in the wrong circumstances. It’s not worth the risk.

For the latter part of the quote, is the transfer itself a problem or do you not have backups? If this is crucial data, any raid solution is not enough. Raid is not a backup, it’s an availability solution.

Probably lack of better knowledge. Truenas is great but it greatly depends on the user understanding all necessary requirements. There is a learning curve.
If you are not the responsible technician, kudos to you for looking into it but it’s probably best if you direct the technician responsible for setting up the server here.


Thanks Chuck. I will have to confer with my tech, as you suggest. It’s certainly possible he is not too familiar with software RAIDs. I know this is the first time he’s used TrueNAS specifically.

To answer your other question, we have installed an HPE LTO-8 external drive for backups and archiving. So it’s certainly doable to send everything to LTO and then restore it from that. I think the biggest thing will be the 24-48 hour downtime it would entail, but that’s certainly a minor factor compared with the possible outcomes of not migrating.

I will echo that “Oh No!”

You will want to change that if you plan to use TrueNAS.

1 Like

This is correct: ZFS expects to have direct access to the drives, which is the reason you have to use IT mode on HBAs.