NVX Memory allocation failed e2fsck: aborted (case 18934057)

I have a ReadyNAS NVX running RAIDiator 4.2.21 where I have been replacing 2TB disks one at a time with 3TB disks. After each 3TB disk is installed, I allow the volume to expand and then move on to the next disk. After the 3rd of 4 disks, I went on vacation and the file system check ran at the normal scheduled time and failed. Now that I am back I have spent quite a bit of time on this and am unable to get the file system to check to pass. It fails at 52%. Here's the end of the log:

Error storing directory block information (inode=91203673, block=0, num=2158820): Memory allocation failed
e2fsck: aborted


Based on various forums, I installed root access and created a /etc/e2fsck.conf file that points to /var/cache/fsck. As it runs, I can see the directory being used:

NAS:/var/cache/fsck# ls -al /var/cache/fsck
total 16
drwxrwxrwx 2 root root 4096 2012-07-07 17:29 .
drwxr-xr-x 11 root root 4096 2012-07-07 17:26 ..
-rw------- 1 root root 4096 2012-07-07 17:29 50f390ac-22de-4de0-b1ea-e3975a394a3b-dirinfo-Xs4m2U
-rw------- 1 root root 4096 2012-07-07 17:29 50f390ac-22de-4de0-b1ea-e3975a394a3b-icount-W9ADMZ



However, it still fails. Based on some reading I wonder whether e2fsck is able to handle such a large volume. My volume configuration is 1xiSCSI 2TB volume and the rest is the normal main volume.

# uname -a
Linux NAS 2.6.37.6.RNx86_32.1.4 #1 Mon May 21 15:49:35 PDT 2012 i686 GNU/Linux

On the bright side, I still have access to my files and am able to write to both volumes, but I am concerned about the failure. I don't know whether anything is truly lost or corrupt. Based on the complete log, it appears that there is enormous amount of corruption that is getting fixed.

So, I suppose the first question is: Although Netgear advertises 3TB support, does it really work w/x86 Linux and 3x3TB drives + 1x2TB drive? Assuming it does, how do I get e2fsck to complete?