Hello. Some of our old projects enable stream caching on their Camel contexts, and set some options that I'm not sure about.
Most of our use cases read data from files, send them to JMS queues before transforming and processing them. Setting a value for "spoolThreshold", if I'm not wrong, forces the memory cache to be written to disk, and thus read again as a string. I don't think it makes sense to get a string from a file via the cache that uses another file. Am I wrong ? For some exchanges, the data size is between 1 and about 250 MB. What should be the right "bufferSize" value. Setting a big value (near the maximum expected size) should reduce IO operation I suppose. And a smaller (or default) value may multiply the buffer operations. Right ? Thanks in advance for your help. Regards.
