Hi, I don't know of a way to completely clear the cache, except the obvious 'close the repository'.
There is a way to reduce the cache size however: ((org.apache.jackrabbit.core.RepositoryImpl)rep). getCacheManager(). setMaxMemory(1); This will however be applied immediately. The cache size is applied once a second at most, and is triggered by CacheManager.cacheAccessed(). Don't forget to change the size back to the old value afterwards, otherwise the repository will be very slow. Regards, Thomas On Tue, Oct 7, 2008 at 11:21 AM, Jacco van Weert <[EMAIL PROTECTED]> wrote: > Hello, > > For some time I am trying to master to memory use of Jackrabbit. > > I am using Jackrabbit in a webserver application (JeCARS). > > When started and accessing nodes there are a couple of megabytes in use, > regardless of the size of the repository itself. > At a certain time I run the "JeCARS internal" garbage collector to remove > expired JCR nodes and also run the Datastore GC. > After it is finished, the memory in use is dramatically higher.... this > because the Datastore checks every node in the repository, but it doesn't > falls back (cache?) > > My question, is there a way to "clear the cache" to reclaim the memory? > Were is memory cached?.... session logout doesn't seem to help. > > > Greetings, > > Jacco > > > > -- > ------------------------------------- > Jacco van Weert -- [EMAIL PROTECTED] > JCR Controller -- http://www.xs4all.nl/~weertj/jcr > JeCARS -- http://jecars.sourceforge.net >
