Thanks for your responses. I narrowed down the faulty data. as per the documentation *Impala timestamp Range:* Allowed date values range from 1400-01-01 to 9999-12-31; I was ingesting data into Kudu from oracle(timestamp in question 06-05-0084 00:00:00) using a spark job and it adds this to kudu table without any problem. and despite the smaller range as per impala documentation I can query this data and it shows -16532:15:03.600000000 with out any errors. I see the error only when I try to insert it from one kudu table into another kudu table through impala sqls.
On Sat, Nov 11, 2017 at 11:11 PM, John Russell <[email protected]> wrote: > Also if the cause of the problem is some sort of out-of-range TIMESTAMP > value, perhaps selecting the MIN() and MAX() of that column would return > the values most likely to be problematic. > > John > > > On Nov 8, 2017, at 8:40 AM, Jim Apple <[email protected]> wrote: > > > > Can you change your query into a SELECT (without UPDATE or CREATE > > TABLE or INSERT, I mean) to find the row in question? > > > > On Tue, Nov 7, 2017 at 11:34 PM, chaitra shivakumar > > <[email protected]> wrote: > >> Hi, > >> > >> We recently upgraded to kudu 1.5 and impala 2.10 on cloudera CDH 5.13 > >> > >> I am trying to merge data from one table into another and using left > outer > >> join. > >> > >> Half way through the merge I get a error which really does not point me > to > >> the problematic row. > >> Error seen is > >> > >> Query Status: Invalid TimestampValue: -16532:15:03.600000000 > >> > >> I can see the error could be because of some conversion issues, but I > was > >> trying to dig a little deeper into the logs to pin point what exactly > >> causes this problem > >> and all I could find was this > >> > >> cc:55 > >> > >> Invalid TimestampValue: -16532:15:03.600000000 > >> @ 0x83d85a impala::Status::Status() > >> @ 0x8415e2 impala::WriteKuduTimestampValue() > >> @ 0x842437 impala::WriteKuduValue() > >> @ 0xce2900 impala::KuduTableSink::Send() > >> @ 0xa50fc4 impala::FragmentInstanceState::ExecInternal() > >> @ 0xa543b9 impala::FragmentInstanceState::Exec() > >> @ 0xa30b38 impala::QueryState::ExecFInstance() > >> @ 0xbd4722 impala::Thread::SuperviseThread() > >> @ 0xbd4e84 boost::detail::thread_data<>::run() > >> @ 0xe6113a (unknown) > >> @ 0x7f56b4210dc5 start_thread > >> @ 0x7f56b3f3dced __clone > >> > >> Data set has a few 100 thousand rows of data so It hard to pinpoint > >> manually. > >> > >> Is there a way to get a more concrete error to pinpoint the problem, or > some > >> resources on how to resolve it. > >> > >> > >> > >> > >> > >> > > -- regards Chaitra
