Issue with sorting in Linux/Unix

Question

I have two very long comma delimited files. Each line has around 101 columns on both the files and each file has 41705 records.

Before Sorting File A & File B both:-

---

01B60D9F4DEB83,FY 
01B60D9F4DEB83,AU
01B60D9F4DEB83,BB
01B60D9F4DEB83,BB
01B60D9F4DEB83,JA

After Sorting File A:-

---

01B60D9F4DEB83,BB
01B60D9F4DEB83,FY
01B60D9F4DEB83,BB
01B60D9F4DEB83,AU
01B60D9F4DEB83,JA

After Sorting File B:-

---

01B60D9F4DEB83,FY 
01B60D9F4DEB83,BB
01B60D9F4DEB83,BB
01B60D9F4DEB83,AU
01B60D9F4DEB83,JA

Command used on bolth Solaris & Linux :-

sort -t"," -k1,1 fileA.csv > FileA_Sort.csv
sort -t"," -k1,1 fileB.csv > FileB_Sort.csv

Can anybody point me why we are getting differences on the same values for the same starting sequence? What am I missing?

Answer 1

you are sorting on first column. When many records have the same value in the first column, they can be in any random order. For big files, the algorithm depends of the available memory (http://vkundeti.blogspot.fr/2008/03/tech-algorithmic-details-of-unix-sort.html).

If you want to preserve the order, you should add the option '-s'.