[Zgemma H7] 5.4 developer builds can't do filesystem check

[BrokenUnusableAccount]

I'm sure at some point with 5.3 images I remember it was possible to do:
Menu / Setup / Storage / Filesystem Check
and choose your box's internal hard drive then
press Green for Check
and an fsck would be done on the hard drive.
This worked even if Timeshift was active.

However now with the 5.4 images It doesn't ever seem to work even if Timeshift is not active.
I always get a message:
umount: Can't unmount /media/hdd: Device or resource busy

 

[bellejt]

same problem in 5.3 . Solution was to warm restart te box

 

[BrokenUnusableAccount]

same problem in 5.3 . Solution was to warm restart te box

When exactly?
I've tried every combination I can think of.
And what exactly do you mean by warm restart?
I'm sure it used to work no problem at some stage.

 

[bellejt]

on my duo2 I sometimes have the same problem.Choose menu - stdby - restart box .Then immediatly after box is restarted do the check.Works everytime here

 

[BrokenUnusableAccount]

Doesn't work for me.
Also "sometimes have the same problem" is not what I'm seeing.
It never works EVER.

 

[BrokenUnusableAccount]

The hard drive in my H7 seems to have a GPT partition table.
Is that normal?


When I try to force an fsck by other means from the command line it seems like perhaps it doesn't like it being GPT.
For instance:

tune2fs -c 1 /dev/sda

says:

tune2fs: Bad magic number in super-block while trying to open /dev/sda
Found a gpt partition table in /dev/sda

 

[BrokenUnusableAccount]

Okay. My bad. I need to do a Linux refresher course.

I needed:

tune2fs -c 1 /dev/sda1

which makes the box do a fsck first thing after every reboot.
That'll do for now.

 

[birdman]

I always get a message:
umount: Can't unmount /media/hdd: Device or resource busy

So what do you have using it?

• Do you have debug logs and write them to that disk?
• Is there a recording running to it?
• Do you have a swap file (that's a Linux swap file) in use there?

 

[BrokenUnusableAccount]

So what do you have using it?

• Do you have debug logs and write them to that disk?
• Is there a recording running to it?
• Do you have a swap file (that's a Linux swap file) in use there?

Ah.
Maybe debug logs are still enabled.

 

[adm]

So what do you have using it?

• Do you have debug logs and write them to that disk?
• Is there a recording running to it?
• Do you have a swap file (that's a Linux swap file) in use there?

Maybe also over the air EPG data being written to HDD

 

[BrokenUnusableAccount]

Maybe also over the air EPG data being written to HDD

As I understand it the EPG is normally only saved to disk just before a reboot and reloaded from disk just after booting.

 

[adm]

As I understand it the EPG is normally only saved to disk just before a reboot and reloaded from disk just after booting.

There are user options to refresh and save at configurable times.

 

[BrokenUnusableAccount]

Ah.
Maybe debug logs are still enabled.

Yep.
That was it.
I'd just assumed that since it was able to check even if Timeshift is active that it could work around logging too.
Wrong!

 

[Joe_90]

Debug logs are normally saved to flash - /home/root/. If you change the default log location to HDD, then the disk will stay active. Not an issue if you are already using permanent timeshift, but it will keep the disk spinning if you are not.