Home | About | Sematext search-lucene.com search-hadoop.com
 Search Hadoop and all its subprojects:

Switch to Plain View
Zookeeper >> mail # user >> question on zookeeper single threaded

Copy link to this message
question on zookeeper single threaded

I am trying to use the 'C' zookeeper APIs from a single threaded application (linking to lib zookeeper_st.so), I noticed that when I call
zoo_create (with ZOO_SEQUENCE), the node gets created but the response does not have the path_buffer set with the path of the actual node created. After a little more digging through, noticed that wait_sync_comletion was a dummy function for single threaded application and in the multithread applicaiotn, it was this function which ensured that the path_buffer was filled up before we return to the caller.

Just wanted to know what needs to be done to make zookeeper_st work properly for single threaded application. Is there a how-to guide? Google search did not give me any interesting results…