g_vfs_done() Read(offsett=* leng=*) ]error=5

[Anitem]

1. which file system is full here?

2 . Then following error keeps scrolling forever.

 

[Anitem]

Thanks in advance.

 

[SirDice]

which file system is full here?

It's mentioned right there, in the error message on the first picture. Twice.

Please don't post pictures of text, they're impossible to quote from.

 

[ralphbsz]

The second one: You are getting IO errors on that disk. Do a "man errno", and you'll see that error 5 is IO error.

To drill down on what is wrong: Look at the dmesg output (which may be in /var/log/messages). You may want to use smartctl to ask the disk whether it is healthy, but the problem could also be the connection to the disk.

 

[cracauer@]

Odd coincidence that you get a filesystem full and I/O errors at the same time. I guess one of the conditions has been on for a while.

 

[SirDice]

Maybe somebody extended the VM's disk image and resized the partition but forgot to growfs(8) the filesystem?

 

[ralphbsz]

Maybe somebody extended the VM's disk image and resized the partition but forgot to growfs(8) the filesystem?

That wouldn't give you IO errors: there would be space on the virtual disk that the file system never accesses. But I like your approach to trying to explain cracauer's coincidence: Maybe someone shrank the virtual disk, then growfs went pear-shaped (*), and now the file system is trying to read blocks that don't exist. Looking at the exist IO error messages in the dmesg output would show that pretty quickly.

(*) Pear-shaped means "not good", which is funny because pears are delicious.

 

[_martin]

It's also interesting why is g_vfs_done() formatting the offset (off_t) to a signed integer.

 

[Anitem]

Maybe somebody extended the VM's disk image and resized the partition but forgot to growfs(8) the filesystem?

No , i haven't done any partition extends as well as resizing.

 

[Anitem]

 

[ralphbsz]

It's also interesting why is g_vfs_done() formatting the offset (off_t) to a signed integer.

Good catch; that's bizarre.

 

[Anitem]

 

[cracauer@]

This system Looks Unhealthy(tm).

Time to copy it to a new filesystem.

 

[SirDice]

What version of FreeBSD is this anyway?

Time to copy it to a new filesystem.

Yeah, that's probably the best solution. You can spend days, even weeks, chasing down the cause while everything's down. Or spend a day setting up a new VM, copy the data and be back up and running. Sometimes you just have to make that decision, costs vs. benefits and all that.

 

[Anitem]

What version of FreeBSD is this anyway?


Yeah, that's probably the best solution. You can spend days, even weeks, chasing down the cause while everything's down. Or spend a day setting up a new VM, copy the data and be back up and running. Sometimes you just have to make that decision, costs vs. benefits and all that.

Free BSD version 11.1

 

[SirDice]

Yeah, best to install a new VM and transfer whatever data you can salvage. FreeBSD 11.1 has been end-of-life since September 2018.

 

[Anitem]

Yeah, best to install a new VM and transfer whatever data you can salvage. FreeBSD 11.1 has been end-of-life since September 2018.

This server is mail server with squirrel mail, Anyway i will try to install newest version, but I am afraid if i can salvage all the data. Thank anyway.