Erroneous 'avgqu-sz' and 'await' values reported in SAR's disk i/o output in Red Hat Enterprise Linux 6

Issue

A Red Hat Enterprise Linux (RHEL) 6 guest running in a Red Hat Enterprise Virtualization (RHEV) 3.0 environment is randomly stalling during the day. Looking at sar data, there are disk I/O values which appear to be invalid.

Sample output from 'sar -d' shows such values in the 'avgqu-sz' and 'await' fields;

00:00:01          DEV       tps  rd_sec/s  wr_sec/s  avgrq-sz  avgqu-sz     await     svctm     %util
10:40:01 PM  dev253-1   3950.86      0.00  32185.57      8.15    491.45    124.39      0.06     23.49
10:50:01 PM  dev253-1   3943.85      0.00  32077.66      8.13   2234.90    566.68      0.11     44.15
11:00:01 PM  dev253-1   3851.51      0.00  31316.39      8.13   1910.65    496.08      0.10     36.65
11:10:01 PM  dev253-1   2778.78      0.00  22765.88      8.19 30599227121898.37    11011745589412.31      0.12      33.95