hi Andy, We'd need to see a more complete log of the verification script run to see what's going on.
If I had to venture a guess if you are working on EC2 you might be running out of /tmp space. The build produces well over a gigabyte of artifacts. If this is the case, we might add an environment variable option to configure where to run the build (something instead of /tmp) - Wes On Mon, Jan 14, 2019 at 8:30 PM Andrew Palumbo <ap....@outlook.com> wrote: > > After following all setup and install steps from: > https://cwiki.apache.org/confluence/display/ARROW/How+to+Verify+Release+Candidates > > > apalumbo@ip-xxx-xxx-xxx-xxx:~/arrow/dev/release$ verify-release-candidate.sh > source 0.12.0 2 > > > > fails with this stack trace: > > > > > ######################################################### > C++ producing, C++ consuming > ########################################################## > ========================================================== > Testing file > /tmp/arrow-0.12.0.5xTVS/apache-arrow-0.12.0/integration/data/simple.json > ========================================================== > -- Creating binary inputs > Traceback (most recent call last): > File "integration_test.py", line 1216, in <module> > run_all_tests(debug=args.debug, tempdir=args.tempdir) > File "integration_test.py", line 1179, in run_all_tests > runner.run() > File "integration_test.py", line 937, in run > self._compare_implementations(producer, consumer) > File "integration_test.py", line 959, in _compare_implementations > producer.json_to_file(json_path, producer_file_path) > File "integration_test.py", line 1086, in json_to_file > return self._run(arrow_path, json_path, 'JSON_TO_ARROW') > File "integration_test.py", line 1080, in _run > run_cmd(cmd) > File "integration_test.py", line 87, in run_cmd > output = subprocess.check_output(cmd, stderr=subprocess.STDOUT) > File > "/tmp/arrow-0.12.0.5xTVS/apache-arrow-0.12.0/test-miniconda/envs/arrow-test/lib/python3.6/subprocess.py", > line 336, in chec > k_output > **kwargs).stdout > File > "/tmp/arrow-0.12.0.5xTVS/apache-arrow-0.12.0/test-miniconda/envs/arrow-test/lib/python3.6/subprocess.py", > line 403, in run > with Popen(*popenargs, **kwargs) as process: > File > "/tmp/arrow-0.12.0.5xTVS/apache-arrow-0.12.0/test-miniconda/envs/arrow-test/lib/python3.6/subprocess.py", > line 709, in __in > it__ > restore_signals, start_new_session) > File > "/tmp/arrow-0.12.0.5xTVS/apache-arrow-0.12.0/test-miniconda/envs/arrow-test/lib/python3.6/subprocess.py", > line 1344, in _ex > ecute_child > raise child_exception_type(errno_num, err_msg, err_filename) > FileNotFoundError: [Errno 2] No such file or directory: > '/tmp/arrow-0.12.0.5xTVS/apache-arrow-0.12.0/cpp/build/release/arrow-json- > integration-test': > '/tmp/arrow-0.12.0.5xTVS/apache-arrow-0.12.0/cpp/build/release/arrow-json-integration-test' > + cleanup > + rm -fr /tmp/arrow-0.12.0.5xTVS > > > > > Jira or is this something obvious that i missed? > > Thank you. > > --Andy > > ________________________________ > From: Andrew Palumbo <andrew_palu...@msn.com> > Sent: Monday, January 14, 2019 9:23 PM > To: dev@arrow.apache.org > Subject: Re: [VOTE] Release Apache Arrow 0.12.0 (RC2) ARROW > > Hello, I've just run the verify scipt on the 0.12.0 tag after configuring an > Ubuntu machine per the instructions. I'm crashing at "c++ producing, c++ > consuming". With a no such file or directory... Arrow-json-integration-test. > > When I was trying to build straight from that submodule there was a > > git submodule update --init command that I had to run. > > Just taking a wild guess that it could be the cause. > > I have to let this message go through to show the stack trace, because I > hadn't confirmed my dev@a.a.o acct.
