rsync to ext2 filesystem issues

[Ramiro Aceves <ea1abz%gmail.com@localhost>]

Hello,

I have an external USB hard disk partitioned with ext2 filesystem 
(/dev/sd0e mounted in /mnt/datos for file sharing between NetBSD and 
Linux) and I have had recurrent problems when rsync-ing files from the 
NetBSD  FFS filesystem to the EXT2 filesystem.

netbsd-nuc$ df -h
Filesystem     Size   Used  Avail %Cap Mounted on
/dev/dk4       192G    94G    89G  52% /
tmpfs          4.0G   844K   4.0G   1% /tmp
kernfs         1.0K   1.0K     0B 100% /kern
ptyfs          1.0K   1.0K     0B 100% /dev/pts
procfs         4.0K   4.0K     0B 100% /proc
tmpfs          4.0G   7.9M   4.0G   1% /var/shm
/dev/sd0e      1.8T   157G   1.5T  10% /mnt/datos

I have this simple rsync script file:

netbsd-nuc$ cat copia_a_mnt_datos.sh
#!/bin/sh
rsync --delete-after -avP /etc /mnt/datos/COPIA_NETBSD_8I7/
rsync --delete-after -avP /usr/pkg/etc /mnt/datos/COPIA_NETBSD_8I7/
rsync --delete-after -avP /root /mnt/datos/COPIA_NETBSD_8I7/
rsync --delete-after -avP --exclude='/home/ramiro/.cache' /home 
/mnt/datos/COPIA_NETBSD_8I7/
rsync --delete-after -avP /var/cron /mnt/datos/COPIA_NETBSD_8I7/
rsync --delete-after -avP /var/log /mnt/datos/COPIA_NETBSD_8I7/
echo $?

After a while rsync-ing many files, it stops for ever and no progression 
is made:

0 (xfr#2350, to-chk=234820/245092)
home/ramiro/.mozilla/firefox/527cpmj0.default-default-2/storage/permanent/chrome/idb/3870112724rsegmnoittet-es.files/2256
          1,028 100%    1.14kB/s    0:00:00 (xfr#2351, 
to-chk=234819/245092)
home/ramiro/.mozilla/firefox/527cpmj0.default-default-2/storage/permanent/chrome/idb/3870112724rsegmnoittet-es.files/2257
          1,028 100%    1.14kB/s    0:00:00 (xfr#2352, 
to-chk=234818/245092)
home/ramiro/.mozilla/firefox/527cpmj0.default-default-2/storage/permanent/chrome/idb/3870112724rsegmnoittet-es.files/2258
000

There are no strange dmesg messages:

..
..
..

[    10.715843] wsmux1: connecting to wsdisplay0
[    10.725843] wskbd1: connecting to wsdisplay0
[    19.875844] wsdisplay0: screen 1 added (default, vt100 emulation)
[    19.875844] wsdisplay0: screen 2 added (default, vt100 emulation)
[    19.875844] wsdisplay0: screen 3 added (default, vt100 emulation)
[    19.875844] wsdisplay0: screen 4 added (default, vt100 emulation)
[    20.665844] audio2: slinear_le:16 2ch 44100Hz, blk 12000 bytes 
(68ms) for recording
netbsd-nuc$



top command for rsync command says:



load averages:  0.54,  0.68,  0.62;               up 0+06:14:33 
           22:18:21
86 processes: 85 sleeping, 1 on CPU
CPU states:  6.5% user,  0.0% nice,  2.5% system,  1.0% interrupt, 89.9% 
idle
Memory: 9992M Act, 448M Inact, 480M Wired, 377M Exec, 7011M File, 3559M Free
Swap: 16G Total, 16G Free / Pools: 1121M Used / Network: 19K In, 31K Out

  PID USERNAME PRI NICE   SIZE   RES STATE       TIME   WCPU    CPU COMMAND
 6639 root      85    0    41M   27M select/7    0:09  0.00%  0.00% 
rsync --delete-after -avP --exclude=/home/ramiro/.cache /home 
/mnt/datos/COPIA_NETBSD_8I7/
 8182 root      82    0    41M 3956K bioloc/0    0:07  0.00%  0.00% 
rsync --delete-after -avP --exclude=/home/ramiro/.cache /home 
/mnt/datos/COPIA_NETBSD_8I7/
 6096 root      85    0    41M   25M select/6    0:03  0.00%  0.00% 
rsync --delete-after -avP --exclude=/home/ramiro/.cache /home 
/mnt/datos/COPIA_NETBSD_8I7/

with systat iostat I do not see any sd0 disk activity at all.

After killing rsync commands I cannot umount sd0e partition. It normally 
ends in having to fsck the partition in the next reboot (about 1 hour of 
fsck)

Why does it stop and stay for ever? Is it a problem with rsync command? 
or is it a problem with ext2 filesystem?

Thanks so much.
Ramiro.