rhbase hb.insert.data.frame error

[该走了]

library(rhbase)
hb.init()
hb.describe.table('testtable')
# Not required to labels to be 'family:label', can just provide 'family'
# which is equivalent to 'family:'
hb.insert("testtable",list(list("20100101",c("x:a","x:f","y","y:w"), list("James Dewey",TRUE, 187.5,189000))))
hb.insert("testtable",list(list("20100102",c("x:a"), list("James Agnew"))))
hb.insert("testtable",list(list("20100103",c("y:a","y:w"), list("Dilbert Ashford",250000))))
hb.insert("testtable",list(list("20100104",c("x:f"), list("Henry Higs"))))


hb.get("testtable",list("20100101","20100102"))
hb.get("testtable",list("20100101","20100102"),c("y")) # columns that start with y
hb.get("testtable",list("20100101","20100102"),c("y:w"))
hb.get("testtable",list("20100101","20100102"),c("y:w","z"))


# delete
hb.delete("testtable","20100103","y:a")
hb.get("testtable","20100103")

## will not get 20100103, since no 'x'
iter <- hb.scan("testtable",startrow="20100100",colspec="x")
while( length(row <- iter$get(1))>0){
  print(row)
}
iter$close()

# scan from beginning
iter <- hb.scan("testtable",startrow="20100100",colspec=c('x','y','z'))
while( length(row <- iter$get(1))>0){
  print(row)
}
iter$close()

# scan uptil 20100103 (not including it)
iter <- hb.scan("testtable",startrow="20100100",end="20100103",colspec=c('x','y'))
while( length(row <- iter$get(1))>0){
  print(row)
}
iter$close()

hb.new.table('testtable','x','y','z')
x <- data.frame(x=c("Winston","Jen","Jasmine"), y=runif(3),z=c(1,2,3))
rownames(x) <- c("20100105","20100106","20100107")
hb.insert.data.frame("testtable",x)

I ran a simple example of rhbase like this, but when I try to instert  a data.frame, I get an error :

Error in hb.insert.data.frame("testtable", x) : 
  rhbase<addAndSendDFBatch>:: Internal error processing mutateRows

Is there any one know how to fix it? It seems some thing wrong with thrift.
thrift 0.9.1
hadoop 2.2.0
hbase 0.96.1