The HC4 is a different form factor, you cannot stack them.
yes hc2 is stackable and it’s great i have two drives .hc2 only has 32 bit max 16 tb hdd. HC4 64 bit and two slots.( hdd , ssd )
1 Like
It is still going strong too! 
I was not able to finish the disk check on it.
No matter with swap on disk or SD card or scratch files on or off.
I think it is time to retire the HC2.
Not necessarily, if you have a more powerful Linux PC or even Windows with wsl2 (you may mount the disk to wsl2: Get started mounting a Linux disk in WSL 2 | Microsoft Learn with --bare option to be able to fix it) and fix it, then use it back on HC2 (do not forget to unmount the disk from wsl2, if you going to use this method).
Thanks, yes, this is what I am currently doing, running this on a vm and trying to repair it.
But if the HC2 is no longer capable to do this on its own then it is no longer a good device to use. I am lucky that I have it here with me, but if you have it at some other place then you are pretty much doomed.
I will surely put the disk back into it if the repair is successful but I will phase them out.
I think it’s a rare case. While my Pi3B+ was alive, I used fsck once there and it didn’t have any issues with a memory consumption during the check. It also has only 1GB of RAM. However, I used 64 bit Ubuntu lite there, not 32 bit Debian.
Very weird.
I know I had used e2fsck on this before.
Probably also a matter of volume size and how full it is. As garbage has accumulated the disk was basically full.
But still weird it should be able to finish with swap added.
The size of the volume perhaps a culprit. My Pi was handling 2TB HDD CMR.
Running for days now and only slow progress but finding lots of unattached inodes.
This looks really bad for the data but also for the 30 days offline period.
Do you have any ideas what I could do?
Unfortunately not. It needs to be fixed.
It is running for 8 days now…
Then I can only suggest to copy all data from this disk somewhere, format it and copy data back.
However, if it would lose more than 4%, it’s the end of this node.
Just to understand: What I have looked up is that this is pass 4 of e2fsck?
So the scan is over and this is the correction of inodes with errors only? It is not going through all inodes of the fs? That means the check could be over any time?
It must not throw any errors and just print stat.
Most of check tools requires to run several times to fix all detected problems.
E2fsck running for 30 days now and still adding inodes to lost and found.
So my node is gone. Great.
Then perhaps you need to format this drive and start over.
However, there is a little chance that the node can survive, if the fsck is close to the finish.
You didn’t receive a DQ message yet, right?
I am pretty sure it has been disqualified by now as it is running 30 days on this computer but before that it was already a week or so that I tried to get fsck to finish on the HC2.
So it is game over for this node.
I’m sorry about that. However, if you didn’t receive a DQ message yet, it can survive.
It is still running e2fsck and I don’t think it will end soon.
Yes, it is really bad. I never thought it could take that long. On Windows with chkdsk maybe, but not Linux. But I have read the e2fsck seems to have some performance issues anyway so that it is not optimized for larger disks. I get that feeling, that software development in general has issues to keep up with disk size evolvement. We will have 30TB disks soon and the idea of running chkdsk or similar on something like that is pure horror.