#IMPLEMENTATION Input was provided by storing four random keys into a char array prior to database population. Using this array in conjunction with a variable that increments and resets to 0 at a threshold of 4, we were able to reference them (in sequence) repeatedly between different searches. #Key Search The key search was done by performing a single query for both binary tree and hash type databases. Hash is expected to return with it’s data in the shortest amount of time. #Data Search The data search is done by setting the cursor to the first position and then iterating it traverses the whole database while finding matches. For the indexfile type, an associated hash database with reversed keys and data entries implements the key search mechanism to retrieve values quicker. Neither database type is expected to do well. #Range Search The range search is done differs for hash and binary tree types; btree uses the DB_SET_RANGE parameter, while hash iterates over the entire database. Indexfile uses the same mechanism as btree to search by range from the primary database. #IndexFile Our indexfile was a hash database that was associated with a btree database. The indexfile featured reversed key/data pairs which allowed for a large increase in the execution time for data searches. This file also allowed the use of the primary database when searching for keys and searching over ranges, as it got the benefits of the btree file system where similar entries reside beside each other in memory.

#Analysis #Key Search The indexfile method uses the primary database for key searches so it’s average result, 49 μs is justifiably identical to btree. Hash is faster than both btree and indexfile, averaging 38 μs. This is due to it’s much faster retrieval of keys as positions inside of an array, rather than having to constantly determine the key’s position relative to a node in a binary tree. #Data Search Indexfile is considerably faster with it’s inverted key/value entries in a hash table, and it computes this task much more efficiently averaging 85 μs. Indexfile is the superior mechanism because it does not have to iterate over the whole database, unlike hash and btree; it implements the key search on secondary database’s keys (representing data). #Range Search Indexfile uses the same mechanism btree uses to search, thus giving it similar results. Btree is most successful because it’s iteration is limited, and not over the entire database. The lower limit is searched for, then the database iterates until hitting the upper limit, preventing the program from checking all 100,000 key/value pairs. The Hash algorithm has to iterate over the whole database file in order to find the similar keys which are distributed randomly throughout the whole file.