Hi,
The add jar/file syntax stated on hive language manual is as below:
ADD { FILE[S] | JAR[S] | ARCHIVE[S] } <filepath1> [<filepath2>]*
I have an application which remotely connects to hiveserver2 over jdbc. It
needs to add local resources before execution of any hive queries. When it
tries to execute the above add jar or add file commands, with the filepath
being from the jdbc client's local filesystem, it gets an error. But when it
gives path local to hiveserver2 filesystem, it works.
What are my options here? One obvious one is to do scp to hiveserver's
filesystem to say /tmp and then do add jar.
The hive 1.2.0 allows ivy urls. I dont think i can use that since my
application generates the files and i don't want to setup ivy repo where i will
need to copy the files before running add jar.
Isn't there any way to specify client's filesystem paths?
-Kiran
