On Tue, Oct 11, 2011 at 1:28 PM, Bob O'Neil <[email protected]> wrote: > I am trying to weigh out the benefits of the cache or no cache setting when > generating the table handling code.
If you don't cache, then every time a request comes in, the load/free functions will be called. For infrequent requests and tables with just a few rows, this is no big deal. Otherwise, you probably want caching. > My content will not come from a physical file that needs to be polled for > changes. Instead, my collection can be written > to by the application, and I would want the changes to be viewable upon the > next read by demand. I suppose I could call > the free() and load() methods directly when the application makes changes to > the collection rather than the timer based updates. That's a bit excessive too. A better idea would be to have the application manipulate the containers directly. Then you never need to free or load after the initial load. If you have really large tables and/or want to eek out every ounce of performance, there are other tricks you can play with too, but they also require more in-depth understanding of the code. The mib2c generated code is meant to be as simple as possible while trying to handle a good bit of the internal details. ------------------------------------------------------------------------------ All the data continuously generated in your IT infrastructure contains a definitive record of customers, application performance, security threats, fraudulent activity and more. Splunk takes this data and makes sense of it. Business sense. IT sense. Common sense. http://p.sf.net/sfu/splunk-d2d-oct _______________________________________________ Net-snmp-coders mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/net-snmp-coders
