Snapshot Differential Algorithm Based On A MD5 Variant

With the rapid development of computer technology and human society, the amout of information continuously grows in a digital geometric trend. Integrating the information effectively will greatly improve production efficiency and bring great benefits to society.The basis and the first step of information integration is incremental detecting. The factors of evaluating an incremental detecting method are the minimum price and efficiency. There are a lot of methods designed specifically for these requirements appeared in these years, such as log analysis, tiggers using methods. But the snapshot differential algorithm, which just simply compares information source's two different snapshots, is in a pasiton of being widely used because of its versatility and simplicity.This paper studies the snapshot differential algorithm which takes an irreplaceable location because of its versatility and portability in varies incremental detection methods. We introduce the traditional SortMerge algorithm and consider the main volume of price generated during comparisons is the multiple IOs. In order to reduce the amount of the IO price of SortMerge algorithm we presented a new algorithm comparing the digested message enhances the efficiency of the algorithm greatly.As the performance of the message digest algorithm implementation is critical to the new modification detection algorithm, a further research in selection and improvement of the digest algorithm is being done. We finally choose the MD5 digest algorithm which is being widely used and made a simplificated variantion and validated the varant. The SortMerge algorithm which based on the MD5 varant gives a better performance than not only the original SortMerge modification detection method and also the SortMerge algorithm which based on the oragnal MD5.