Dead HDD Replacement In Hexos

[ubergeek]

Is there a standard way to replace a drive in HexOs yet ? I know in Truenas I can out the drive and replace it in the pool but I am unsure if there is way it should be done in Hex. Yes I know we are still in beta mode and this may not have been addressed yet but figured I would ask the question 

[Mawson]

Very curious about this also.

What shows up if you click "replace" ?

[ubergeek]

11 minutes ago, Mawson said:

Very curious about this also.

What shows up if you click "replace" ?

It shows the spair for pool 2 not pool 1 where the drive is dead. I suspect because the hot spair was used already from pool 1. 

 

I would love to see the option to add more drives and allow them to be assigned hot spair for X pool 

[ubergeek]

So a bit if a update, I have been fighting this all week, I isolates the dead drive in truenas , stopped and outted the drive . allowed the system to resilver then put in a new drive to make a hot spare and everything was running smooth. The next day is when all the goofy stuff was going on with the deck so I cam home and started making back ups for truenas ect. went to bed and everything was working well. While at work the following day I started getting notifications from Up time robot telling me that the links going to my hexos was down . Logged into the deck and server not online. After some trouble shooting I had found that I no longer had tailscale access nor could i ping it from any internal computer. Got home and got into the truenas interface and rebooted the server. Once I gained UI access it showed it was trying to resilver the pool again, not thinking about I let it ride out. Well it was stuck at 18% silvering for 2 days so I started to try and delete the array and start over "just move all my data back." AND now I am at a point where im just going to rebuild the OS and all because the only access that I have is Truenas interface

[Mawson]

Interesting...
@jonp, this seems like something that could use further digging into...

[Dylan]

On 1/17/2025 at 7:09 PM, ubergeek said:

So a bit if a update, I have been fighting this all week, I isolates the dead drive in truenas , stopped and outted the drive . allowed the system to resilver then put in a new drive to make a hot spare and everything was running smooth. The next day is when all the goofy stuff was going on with the deck so I cam home and started making back ups for truenas ect. went to bed and everything was working well. While at work the following day I started getting notifications from Up time robot telling me that the links going to my hexos was down . Logged into the deck and server not online. After some trouble shooting I had found that I no longer had tailscale access nor could i ping it from any internal computer. Got home and got into the truenas interface and rebooted the server. Once I gained UI access it showed it was trying to resilver the pool again, not thinking about I let it ride out. Well it was stuck at 18% silvering for 2 days so I started to try and delete the array and start over "just move all my data back." AND now I am at a point where im just going to rebuild the OS and all because the only access that I have is Truenas interface

Any update/change with this?

[ScottC.]

Looking at this from the HexOS side - I have a 2-disk mirror where one of the drives has failed and the other is warning (let this be a lesson that similar drives brought into service at the same time have a tendency to die around the same time...assuming they survive infant death periods).

When I click "Replace", I get this nice popup that says to "Select a replacement from one of your currently matching unused drives below...". Only I don't have any unused drives and none that match the failed one in the system, used or otherwise.  I have a perfect replacement sitting here on my desk.  What do?

[ScottC.]

4 hours ago, ScottC. said:

Looking at this from the HexOS side - I have a 2-disk mirror where one of the drives has failed and the other is warning (let this be a lesson that similar drives brought into service at the same time have a tendency to die around the same time...assuming they survive infant death periods).

When I click "Replace", I get this nice popup that says to "Select a replacement from one of your currently matching unused drives below...". Only I don't have any unused drives and none that match the failed one in the system, used or otherwise.  I have a perfect replacement sitting here on my desk.  What do?

Google and the Truenas admin interface for the win.

[deverett85]

I've just had a drive report smart errors and had to shut down the machine to pull to see if it was one covered by a warranty. When I got back in after checking the drives now fully dead and the pools reported itself as unhealthy.

Problem in HexOS was the dead drive vanished completely from the UI so all I could do was expand the pool again which isn't the right thing to do.

Resolved it with a dive in to the truenas UI to see the dead drive and hit replace with the new one but the Hex UI really needs to keep the dead/removed drive in the pool so the replace button is there.