I have a very large file (millions of records) that I will need to store as a db file. We will only do look up operations on the file so far. However, I would like the following data access pattern First, I'm seeking to a specific location in the file and then reading sequentially a number of records; then, again, I may want to jump to another specific location in the file and reading sequentially a number of records. The data will be stored in the db in sorted order. I know that BDB B-trees fundamentally store keys in sorted order and a value associated with that key. However, using BTree will give me O(log N) each time when finding a specific location. Is there a way that I can combine hash and BTree to improve the performance. I.g., using hash index to find the specific location in O(1) and then do sequential reads (this can only be done by using BTree). So, I am wondering if I can accomplish this task in BDB? Thank you so much for any suggestions!!!
Since you mention the hash access method, I assume you're posting to the wrong forum. This is the forum for BDB JE, which only supports the Btree access method. The BDB product (C-based product with Java JNI API) has a hash access method, so you should probably re-post your question to the BDB forum: