Millions of “shrink_slab: ext4_es_scan+0x0/0x130 [ext4] negative objects to delete nr=-2147483358”

On two machines we have a reoccurring problem where the root filesystem gets 100% full because of /var/log/messages filling up with millions of messages as shown below.

We have many other machines with the identical HW/SW configuration, but only two are showing this problem. Of course, when the root file system is 100% full other things don't go all that well.

I've searched around for what causes it but came up empty. Any ideas or suggestions? I've never seen this problem on any other machines...

There's a full list of error messages here:
https://serverfault.com/questions/99...bjects-to-dele

Dec 8 14:43:12 localhost kernel: vmscan: shrink_slab: ext4_es_scan+0x0/0x130 [ext4] negative objects to delete nr=-2147483134
Dec 8 14:43:12 localhost kernel: vmscan: shrink_slab: ext4_es_scan+0x0/0x130 [ext4] negative objects to delete nr=-2147483166
Dec 8 14:43:12 localhost kernel: vmscan: shrink_slab: ext4_es_scan+0x0/0x130 [ext4] negative objects to delete nr=-2147483198
Dec 8 14:43:12 localhost kernel: vmscan: shrink_slab: ext4_es_scan+0x0/0x130 [ext4] negative objects to delete nr=-2147483230
Dec 8 14:43:12 localhost kernel: vmscan: shrink_slab: ext4_es_scan+0x0/0x130 [ext4] negative objects to delete nr=-2147483262