How Can Row-Wise Compare Say Table Is Identical but Bulk Checksum Say It Is Different?

Issue

Row-by-Row Compare say a table is identical, but Bulk Compare using checksums say it's different.

Environment

HVR 5

Resolution

There are various reasons why bulk compare could say a table is different, but row by row compare says it's the same.

• 'Noise'. Unused bytes in the HVR transport format could cause a different checksum, even though the values are identical.
• Float data types are sometimes 'lossy'. Is 0 the same as 1E-200? Row by row compare uses a matching algorithm with a tolerance for float rounding inaccuracies.
• Other coercion errors.

Do the following to troubleshoot such problems.

When such a problem is detected it is important to get a specific test case back to HVR technical support.

1. First, make a channel with only the specific table.
2. 'Chop' it down to the key columns and a column with the false difference. Columns can be chopped either by removing them in the GUI or by adding ColumnProperties/IgnoreDuringCompare.
3. Keep doing that until the number of columns has been minimized.
4. Finally, chop down the number of rows. This can be done be defining action {Group=* Table=tab1 Restrict /RefreshCondition="{k} <= {hvr_var_k_min} and {hvr} >= {hvr_var_k_max}" /Context=chop}.
5. Use HVR Compare with the 'chop' context enabled (in the Contexts tab) and experiment with min and max values until the difference is localized.
6. Send a dump of the bad rows and the channel definition to HVR Technical Support.

Cause

Conceptually, the bulk compare looks at the raw bytes as HVR transports them over the network; it just checks if they are the same. The row-wise compare unpacks these bytes into the correct datatype and then compares them; it checks which value is bigger (not just if they are the same).

There could be a difference between row-wise and bulk compare, but this would be an HVR bug. The problem would be that bulk compare says a table is different, whereas row-by-row compare says it is the same. The opposite problem (bulk compare says same, whereas row-by-row compare says different) has not been seen before and would be weird.