Files
impala/tests/shell/util.py
Tim Armstrong f1f3ae9ec2 IMPALA-7290: part 2: Add HS2 support to Impala shell
HS2 is added as an option via --protocol=hs2. The user-visible
differences in behaviour are minimal. Beeswax is still the
default and can be explicitly enabled via --protocol=beeswax
but will be deprecated. The default is unchanged because
changing the default could break certain workflows, e.g.
those that explicitly specify the port with -i or deployments
that hit --fe_service_threads for HS2 and somehow rely on
impala-shell not contributing to that limit. For most
workflows the change is transparent and we should change
the default in a major version change.

This support requires Impala-specific extensions to
the HS2 interface, similar to the existing extensions
to Beeswax. Thus the HS2 shell is only
forwards-compatible with newer Impala versions.
I considered trying to gracefully degrade when the
new extensions weren't present, but it didn't seem to be
worth the ongoing testing effort.

Differences between HS2 and Beeswax are abstracted into
ImpalaClient subclasses.
Here are the changes required to make it work:
* Switch to TBinaryProtocolAccelerated to avoid perf
  regression. The HS2 protocol requires decoding
  more primitive values (because its not a string-per-row),
  which was slow with the pure python implementation of
  TBinaryProtocol.
* Added bitarray module to efficiently unpack null indicators
* Minimise invasiveness of changes by transposing and stringifying
  the columnar results into rows in impala_client.py. The transposition
  needs to happen before display anyway.
* Add PingImpalaHS2Service() to get back version string and webserver
  address.
* Add CloseImpalaOperation() extension to return DML row counts. This
  possibly addresses IMPALA-1789, although we need to confirm that
  this is a sufficient solution.
* Add is_closed member to query handles to avoid shell independently
  tracking whether the query handle was closed or not.
* Include query status in HS2 log to match beeswax.
* HS2 GetLog() command now includes query status error message for
  consistency with beeswax.
* "set"/"set all" uses the client requests options, not the session
  default. This captures the effective value of TIMEZONE, which
  was previously missing. This also requires test changes where
  the tests set non-default values, e.g. for ABORT_ON_ERROR.
* "set all" on the server side returns REMOVED query options - the
  shell needs to know these so it can correctly ignore them.
* Clean up self.orig_cmd/self.last_leading comment argument
  passing to avoid implicit parameter passing through multiple
  function calls.
* Clean up argument handling in shell tests to consistently pass
  around lists of arguments instead of strings that are subject
  to shell tokenisation rules.
* Consistently close connections in the shell to avoid leaking
  HS2 sessions. This is enforced by making ImpalaShell a context
  manager and also eliminating all sys.exit() calls that would
  bypass the explicit connection closing.

Testing:
* Shell tests can run with both protocols
* Add tests for formatting of all types and NULL values
* Added testing for floating point output formatting, which does
  change as a result of switching to server-side vs client-side
  formatting.
* Verified that newly-added tests were actually going through HS2
  by disabling hs2 on the minicluster and running tests.
* Add checks to test_verify_metrics.py to ensure that no sessions
  are left open at the end of tests.

Performance:
Baseline from beeswax shell for large extract is as follows:

  $ time impala-shell.sh -B -q 'select * from tpch_parquet.orders' > /dev/null
  real    0m6.708s
  user    0m5.132s
  sys     0m0.204s

After this change it is somewhat slower, but we generally don't consider
bulk extract performance through the shell to be perf-critical:
  real    0m7.625s
  user    0m6.436s
  sys     0m0.256s

Change-Id: I6d5cc83d545aacc659523f29b1d6feed672e2a12
Reviewed-on: http://gerrit.cloudera.org:8080/12884
Reviewed-by: Impala Public Jenkins <impala-public-jenkins@cloudera.com>
Tested-by: Impala Public Jenkins <impala-public-jenkins@cloudera.com>
2019-06-20 10:23:28 +00:00

238 lines
10 KiB
Python
Executable File

#!/usr/bin/env impala-python
# encoding=utf-8
#
# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements. See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership. The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License. You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied. See the License for the
# specific language governing permissions and limitations
# under the License.
import os
import pytest
import re
import shlex
import time
from subprocess import Popen, PIPE
from tests.common.environ import (IMPALA_LOCAL_BUILD_VERSION,
IMPALA_TEST_CLUSTER_PROPERTIES)
from tests.common.impala_test_suite import (IMPALAD_BEESWAX_HOST_PORT,
IMPALAD_HS2_HOST_PORT)
SHELL_HISTORY_FILE = os.path.expanduser("~/.impalahistory")
IMPALA_HOME = os.environ['IMPALA_HOME']
if IMPALA_TEST_CLUSTER_PROPERTIES.is_remote_cluster():
# With remote cluster testing, we cannot assume that the shell was built locally.
IMPALA_SHELL_EXECUTABLE = os.path.join(IMPALA_HOME, "bin/impala-shell.sh")
else:
# Test the locally built shell distribution.
IMPALA_SHELL_EXECUTABLE = os.path.join(
IMPALA_HOME, "shell/build", "impala-shell-" + IMPALA_LOCAL_BUILD_VERSION,
"impala-shell")
def assert_var_substitution(result):
assert_pattern(r'\bfoo_number=.*$', 'foo_number= 123123', result.stdout, \
'Numeric values not replaced correctly')
assert_pattern(r'\bfoo_string=.*$', 'foo_string=123', result.stdout, \
'String values not replaced correctly')
assert_pattern(r'\bVariables:[\s\n]*BAR:\s*[0-9]*\n\s*FOO:\s*[0-9]*', \
'Variables:\n\tBAR: 456\n\tFOO: 123', result.stdout, \
"Set variable not listed correctly by the first SET command")
assert_pattern(r'\bError: Unknown variable FOO1$', \
'Error: Unknown variable FOO1', result.stderr, \
'Missing variable FOO1 not reported correctly')
assert_pattern(r'\bmulti_test=.*$', 'multi_test=456_123_456_123', \
result.stdout, 'Multiple replaces not working correctly')
assert_pattern(r'\bError:\s*Unknown\s*substitution\s*syntax\s*' +
r'\(RANDOM_NAME\). Use \${VAR:var_name}', \
'Error: Unknown substitution syntax (RANDOM_NAME). Use ${VAR:var_name}', \
result.stderr, "Invalid variable reference")
assert_pattern(r'"This should be not replaced: \${VAR:foo} \${HIVEVAR:bar}"',
'"This should be not replaced: ${VAR:foo} ${HIVEVAR:bar}"', \
result.stdout, "Variable escaping not working")
assert_pattern(r'\bVariable MYVAR set to.*$', 'Variable MYVAR set to foo123',
result.stderr, 'No evidence of MYVAR variable being set.')
assert_pattern(r'\bVariables:[\s\n]*BAR:.*[\s\n]*FOO:.*[\s\n]*MYVAR:.*$',
'Variables:\n\tBAR: 456\n\tFOO: 123\n\tMYVAR: foo123', result.stdout,
'Set variables not listed correctly by the second SET command')
assert_pattern(r'\bUnsetting variable FOO$', 'Unsetting variable FOO',
result.stdout, 'No evidence of variable FOO being unset')
assert_pattern(r'\bUnsetting variable BAR$', 'Unsetting variable BAR',
result.stdout, 'No evidence of variable BAR being unset')
assert_pattern(r'\bVariables:[\s\n]*No variables defined\.$', \
'Variables:\n\tNo variables defined.', result.stdout, \
'Unset variables incorrectly listed by third SET command.')
assert_pattern(r'\bNo variable called NONEXISTENT is set', \
'No variable called NONEXISTENT is set', result.stdout, \
'Problem unsetting non-existent variable.')
assert_pattern(r'\bVariable COMMENT_TYPE1 set to.*$',
'Variable COMMENT_TYPE1 set to ok', result.stderr,
'No evidence of COMMENT_TYPE1 variable being set.')
assert_pattern(r'\bVariable COMMENT_TYPE2 set to.*$',
'Variable COMMENT_TYPE2 set to ok', result.stderr,
'No evidence of COMMENT_TYPE2 variable being set.')
assert_pattern(r'\bVariable COMMENT_TYPE3 set to.*$',
'Variable COMMENT_TYPE3 set to ok', result.stderr,
'No evidence of COMMENT_TYPE3 variable being set.')
assert_pattern(r'\bVariables:[\s\n]*COMMENT_TYPE1:.*[\s\n]*' + \
'COMMENT_TYPE2:.*[\s\n]*COMMENT_TYPE3:.*$',
'Variables:\n\tCOMMENT_TYPE1: ok\n\tCOMMENT_TYPE2: ok\n\tCOMMENT_TYPE3: ok', \
result.stdout, 'Set variables not listed correctly by the SET command')
def assert_pattern(pattern, result, text, message):
"""Asserts that the pattern, when applied to text, returns the expected result"""
m = re.search(pattern, text, re.MULTILINE)
assert m and m.group(0) == result, message
def run_impala_shell_cmd(vector, shell_args, env=None, expect_success=True,
stdin_input=None, wait_until_connected=True):
"""Runs the Impala shell on the commandline.
'shell_args' is a string which represents the commandline options.
Returns a ImpalaShellResult.
"""
result = run_impala_shell_cmd_no_expect(vector, shell_args, env, stdin_input,
expect_success and wait_until_connected)
if expect_success:
assert result.rc == 0, "Cmd %s was expected to succeed: %s" % (shell_args,
result.stderr)
else:
assert result.rc != 0, "Cmd %s was expected to fail" % shell_args
return result
def run_impala_shell_cmd_no_expect(vector, shell_args, env=None, stdin_input=None,
wait_until_connected=True):
"""Runs the Impala shell on the commandline.
'shell_args' is a string which represents the commandline options.
Returns a ImpalaShellResult.
Does not assert based on success or failure of command.
"""
p = ImpalaShell(vector, shell_args, env=env, wait_until_connected=wait_until_connected)
result = p.get_result(stdin_input)
return result
def get_impalad_host_port(vector):
"""Get host and port to connect to based on test vector provided."""
protocol = vector.get_value("protocol")
if protocol == 'hs2':
return IMPALAD_HS2_HOST_PORT
else:
assert protocol == 'beeswax', protocol
return IMPALAD_BEESWAX_HOST_PORT
def get_impalad_port(vector):
"""Get integer port to connect to based on test vector provided."""
return int(get_impalad_host_port(vector).split(":")[1])
def get_shell_cmd(vector):
"""Get the basic shell command to start the shell, given the provided test vector.
Returns the command as a list of string arguments."""
return [IMPALA_SHELL_EXECUTABLE, "--protocol={0}".format(vector.get_value("protocol")),
"-i{0}".format(get_impalad_host_port(vector))]
def get_open_sessions_metric(vector):
"""Get the name of the vector that tracks open sessions for the protocol in vector."""
protocol = vector.get_value("protocol")
if protocol == 'hs2':
return 'impala-server.num-open-hiveserver2-sessions'
else:
assert protocol == 'beeswax', protocol
return 'impala-server.num-open-beeswax-sessions'
class ImpalaShellResult(object):
def __init__(self):
self.rc = 0
self.stdout = str()
self.stderr = str()
class ImpalaShell(object):
"""A single instance of the Impala shell. The proces is started when this object is
constructed, and then users should repeatedly call send_cmd(), followed eventually by
get_result() to retrieve the process output. This constructor will wait until
Impala shell is connected for the specified timeout unless wait_until_connected is
set to False or --quiet is passed into the args."""
def __init__(self, vector, args=None, env=None, wait_until_connected=True, timeout=60):
self.shell_process = self._start_new_shell_process(vector, args, env=env)
# When --quiet option is passed to Impala shell, we should not wait until we see
# "Connected to" because it will never be printed to stderr.
if wait_until_connected and (args is None or "--quiet" not in args):
start_time = time.time()
connected = False
while time.time() - start_time < timeout and not connected:
connected = "Connected to" in self.shell_process.stderr.readline()
assert connected, "Impala shell is not connected"
def pid(self):
return self.shell_process.pid
def send_cmd(self, cmd):
"""Send a single command to the shell. This method adds the end-of-query
terminator (';'). """
self.shell_process.stdin.write("%s;\n" % cmd)
self.shell_process.stdin.flush()
# Allow fluent-style chaining of commands
return self
def wait_for_query_start(self):
"""If this shell was started with the '-q' option, this mathod will block until the
query has started running"""
# readline() will block until a line is actually printed, so this loop should always
# read something like:
# Starting Impala Shell without Kerberos authentication
# Connected to localhost:21000
# Server version: impalad version...
# Query: select sleep(10)
# Query submitted at:...
# Query progress can be monitored at:...
# We stop at 10 iterations to prevent an infinite loop if somehting goes wrong.
iters = 0
while "Query progress" not in self.shell_process.stderr.readline() and iters < 10:
iters += 1
def get_result(self, stdin_input=None):
"""Returns an ImpalaShellResult produced by the shell process on exit. After this
method returns, send_cmd() no longer has any effect."""
result = ImpalaShellResult()
result.stdout, result.stderr = self.shell_process.communicate(input=stdin_input)
# We need to close STDIN if we gave it an input, in order to send an EOF that will
# allow the subprocess to exit.
if stdin_input is not None: self.shell_process.stdin.close()
result.rc = self.shell_process.returncode
return result
def _start_new_shell_process(self, vector, args=None, env=None):
"""Starts a shell process and returns the process handle"""
cmd = get_shell_cmd(vector)
if args is not None: cmd += args
if not env: env = os.environ
# Don't inherit PYTHONPATH - the shell launch script should set up PYTHONPATH
# to include dependencies. Copy 'env' to avoid mutating argument or os.environ.
env = dict(env)
if "PYTHONPATH" in env:
del env["PYTHONPATH"]
return Popen(cmd, shell=False, stdout=PIPE, stdin=PIPE, stderr=PIPE,
env=env)