hi, I would like to elaborate on what I wish to do. I want to store the results of hive query in a database instead of just displaying it on the hive shell. For this I need to modify/ add code at two places. 1. To fetch the schema from metastore, so that i can create a table in mySql database 2. Get a handle of a cursor, so that i can dump the result into the database.
I am unable to figure out how to write my code by implementing an interface and create jar file so that I can see the results when I run hive again. I have imported Hive src folder to eclipse, but I am not sure what should I do next. Your help would be greatly appreciated. regards, Abhinav Narain On Fri, Jan 21, 2011 at 2:30 PM, abhinav narain <abhinavnarai...@gmail.com>wrote: > Thanks for the prompt replies. > I have gone through the code. > But i am unable to understand how can I integrate my own class and compile > the code in a jar to be executed later. > I am undergraduate and have not worked on such big project. > > Some hints will be very helpful. > > thanks again for the replies. > > regards, > Abhinav Narain > > On Thu, Jan 20, 2011 at 1:02 PM, Carl Steinbach <c...@cloudera.com> wrote: > >> Hi Abhinav, >> >> 1. >> > How does one have access to metastore of Hive, to retrieve the schema >> > information from it. >> > I cant find a file or other such thing in source code. >> > >> >> Take a look at the IMetaStoreClient interface and HiveMetaStoreClient >> class, >> but bear in mind that the fundamental definition of the MetaStore >> interface >> is contained in the Thrift IDL file located >> here: metastore/if/hive_metastore.thrift. IMetaStoreClient actually >> defines >> a wrapper interface around the code generated by the Thrift compiler based >> on the definitions in hive_metastore.thrift >> >> You can also find some good code examples in TestHiveMetaStore. >> >> >> > 2. >> > How does one get the handle of the result that is produced after the >> query. >> > >> > If someone can tell about where to look for the answers, that will also >> > help >> > >> >> Here are the relevant pieces of code that you should look at: >> >> service/if/hive_service.thrift >> service/src/java/org/apache/hadoop/hive/service/HiveServer.java >> service/src/test/org/apache/hadoop/hive/service/TestHiveServer.java >> >> The interface for executing queries and fetching results is defined in >> hive_service.thrift and consists of the following methods: >> >> void execute(string query) >> string fetchOne() >> list<string> fetchN(i32 numRows) >> list<string> fetchAll() >> >> Since execute() does not return a query ID the Thrift client is limited to >> executing/fetching the results of a single query at a time. >> >> Hope this helps. >> >> Carl >> > >
