CODEWITHSUNDEEP
javastoragehashtable
Kaleb Blue
Kaleb Blue

Reputation: 507

Storing in Hashtables

I an working on an application that might potentially get thousands and thousands of messages (perhaps millions). And I want to store these messages in a hashtable for easy lookup since each message has an id. Is this a good idea? If not, what's the best data structure or way to go about this. Thank you.

Upvotes: 0

Views: 58

Answers (2)

Vasu
Vasu

Reputation: 22452

Using Map (data will be stored main memory) is simple, but should be the least preferable and non realistic option, as you need to and implement/reinvent the logic for the data expiration, clustering, etc.. by yourself.

Using Caching frameworks (data will be stored main memory), this can be chosen only if you have an idea about how much size of data and how long the data needs to be resided in the cache (i.e., when the data can expired and removed), this option limits the data size to the max size of the JVM Heap space.

Using Database (data will be stored in disc space), this is the ideal option for holding millions of data, but comes with a cost as disc operations takes more time compared to the in memory operations.

Upvotes: 0

Stephen C
Stephen C

Reputation: 719541

Is this a good idea?

Keeping an unbounded amount of data in an in-memory data structure is a bad idea. You will eventually run out of memory, and your application will crash.

If you are able to discard old "messages" so that you can place a reasonable bound on the amount of memory the application needs, then this could be a viable solution.

However, as the comments point out there are other solutions (distibuted memory caches, SQL databases, NoSQL databases, etcetera) that could well be better, depending on how much data there is and how fast access really needs to be.

Upvotes: 1

Related Questions