Resource List Of Performance Issues For Java Developers
July 29, 2010
As a resource for the aspiring java developer, here is a list of problems related to application performance.
1. The no.1 problem is far too many database calls per transaction: This refers to requesting more data than is required for a particular transaction. In addition, the exact data may be requested more than once. And the third type of problem that leads to a high number of data requests is making multiple queries to obtain specific data.
2. Oversynchronization: Although synchronization is vital in protecting shared data,developers often resort to synchronizing extensive code sequences. Although this is not a problem for low load on local developer workstation, in a performance environment this can lead to performance and scalability issues.
3. Most libraries provide easy remote communication. For instance, there is not much difference for the developer whether the request is local or remote. Hence, there is very little understanding of what occurs under the remoting hood and this leads to problems with serialization, memory usage, latency and network traffic that is linked with each remoting call. This again ends in performance and scalability problems.
4. In addition, managed runtime environtiments like .NET and Java have the benefit of assisting with the management of memory by providing garbage collectors (GC). However, a GC will not prevent leakage of memory. Ultimately, this will lead to a leak in memory that could result in Out of Memory Exception. In this regard, it is important to release object references right after they are completed.
Comments
Got something to say?
-
Advertisements
Text Links
