I have to catch a DUT before it gets reimaged.

This happens also on machines other than samus. Any ideas? Notice these directories are part of the system image.

ssh root@chromeos2-row1-rack7-host2.cros

7995.0.0 (Official Build) dev-channel samus test

localhost modules # ls
ls: cannot access gles3: Input/output error
egl  gles2  gles3  gles31
localhost modules # cd gles3
-bash: cd: gles3: Input/output error

grep -i ext4 /var/log/messages
2016-03-01T23:15:41.551082-08:00 CRIT kernel: [    2.564418] EXT4-fs error (device sda1): ext4_mb_generate_buddy:756: group 5, 18799 clusters in bitmap, 18780 in gd; block bitmap corrupt.
2016-03-01T23:20:41.139974-08:00 NOTICE kernel: [  302.361795] EXT4-fs (sda1): error count: 17
2016-03-01T23:20:41.139997-08:00 NOTICE kernel: [  302.361824] EXT4-fs (sda1): initial error at 1456901437: ext4_mb_generate_buddy:756
2016-03-01T23:20:41.140001-08:00 NOTICE kernel: [  302.361843] EXT4-fs (sda1): last error at 1456902941: ext4_mb_generate_buddy:756
2016-03-01T23:25:42.711851-08:00 CRIT kernel: [  604.236020] EXT4-fs error (device sda1): ext4_lookup:1436: inode #1570531: comm ls: deleted inode referenced: 1570693
2016-03-01T23:25:47.477857-08:00 CRIT kernel: [  609.006494] EXT4-fs error (device sda1): ext4_lookup:1436: inode #1570531: comm bash: deleted inode referenced: 1570693
2016-03-01T23:25:47.510867-08:00 CRIT kernel: [  609.039332] EXT4-fs error (device sda1): ext4_lookup:1436: inode #1570531: comm bash: deleted inode referenced: 1570693
2016-03-01T23:25:47.510896-08:00 CRIT kernel: [  609.039460] EXT4-fs error (device sda1): ext4_lookup:1436: inode #1570531: comm bash: deleted inode referenced: 1570693

Only changes I'm aware of would be if these DUTs recently moved to the 2.3 SSD firmware. But the unit from comment #2 is on dev-channel, so it should have gotten the update months ago.

Gwendal, any ideas?

Activating. Please assign to the right owner and the appropriate priority.

Probably fixed right now.