> On Jan. 29, 2017, 5:09 p.m., Nate Cole wrote: > > ambari-common/src/main/python/ambari_commons/shell.py, lines 141-142 > > <https://reviews.apache.org/r/56059/diff/1/?file=1618458#file1618458line141> > > > > Seems awfully specific. Would be more flexible if we could pass in the > > kill delay as an argument whose default is 5 (the graceful_kill_delay). > > Dmitro Lisnichenko wrote: > Do you mean having a single graceful_kill_delay parameter (the time > between SIGTERM and SIGKILL)? Or you mean exposing > yum_rpm_gracefull_kill_delay parameter for timeout between sending SIGTERM to > sensitive process like yum, and sending SIGKILL?
What I mean is that yumrpm.py is the one building the command to run. I would prefer supplying the timeout there instead of using the yum_rpm_present as it's just something else we have to remember that we did 6 months from now. We already know when we're calling yum. We have a ton of arguments to control the timeout behavior, can we just be using them? Can we use try_sleep, timeout, on_timeout, timeout_kill_strategy etc to our advantage (see shell.py/checked_call for complete list). - Nate ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/56059/#review163443 ----------------------------------------------------------- On Jan. 29, 2017, 7:27 a.m., Dmitro Lisnichenko wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/56059/ > ----------------------------------------------------------- > > (Updated Jan. 29, 2017, 7:27 a.m.) > > > Review request for Ambari, Alejandro Fernandez, Dmytro Grinenko, Jonathan > Hurley, and Nate Cole. > > > Bugs: AMBARI-19768 > https://issues.apache.org/jira/browse/AMBARI-19768 > > > Repository: ambari > > > Description > ------- > > *Steps* > # Deploy HDP-2.5.0.0 with Ambari 2.4.1.0 > # Upgrade ambari to 2.5.0.0-481 (I did not register Falcon library, as the > jar was already present in /var/lib/ambari-server/resources/je-5.0.73.jar on > Ambari server node) > # Register HDP-2.6.0.0-216 > # Start package installation > > *Result:* > Got below errors: > {code} > 2016-12-16 13:47:10,419|INFO|MainThread|machine.py:145 - > run()|CRITICAL:yum.main: > 2016-12-16 13:47:10,419|INFO|MainThread|machine.py:145 - run()| > 2016-12-16 13:47:10,419|INFO|MainThread|machine.py:145 - run()|Error: rpmdb > open failed > 2016-12-16 13:47:10,420|INFO|MainThread|machine.py:145 - run()|Traceback > (most recent call last): > 2016-12-16 13:47:10,420|INFO|MainThread|machine.py:145 - run()|File > "/var/lib/ambari-agent/cache/custom_actions/scripts/install_packages.py", > line 166, in actionexecute > 2016-12-16 13:47:10,420|INFO|MainThread|machine.py:145 - run()|ret_code = > self.install_packages(package_list) > 2016-12-16 13:47:10,420|INFO|MainThread|machine.py:145 - run()|File > "/var/lib/ambari-agent/cache/custom_actions/scripts/install_packages.py", > line 400, in install_packages > 2016-12-16 13:47:10,420|INFO|MainThread|machine.py:145 - run()|if not > verifyDependencies(): > 2016-12-16 13:47:10,421|INFO|MainThread|machine.py:145 - run()|File > "/usr/lib/python2.6/site-packages/resource_management/libraries/functions/packages_analyzer.py", > line 311, in verifyDependencies > 2016-12-16 13:47:10,421|INFO|MainThread|machine.py:145 - run()|code, out = > rmf_shell.checked_call(cmd, sudo=True) > 2016-12-16 13:47:10,421|INFO|MainThread|machine.py:145 - run()|File > "/usr/lib/python2.6/site-packages/resource_management/core/shell.py", line > 72, in inner > 2016-12-16 13:47:10,421|INFO|MainThread|machine.py:145 - run()|result = > function(command, **kwargs) > 2016-12-16 13:47:10,421|INFO|MainThread|machine.py:145 - run()|File > "/usr/lib/python2.6/site-packages/resource_management/core/shell.py", line > 102, in checked_call > 2016-12-16 13:47:10,422|INFO|MainThread|machine.py:145 - run()|tries=tries, > try_sleep=try_sleep, timeout_kill_strategy=timeout_kill_strategy) > 2016-12-16 13:47:10,422|INFO|MainThread|machine.py:145 - run()|File > "/usr/lib/python2.6/site-packages/resource_management/core/shell.py", line > 150, in _call_wrapper > 2016-12-16 13:47:10,422|INFO|MainThread|machine.py:145 - run()|result = > _call(command, **kwargs_copy) > 2016-12-16 13:47:10,422|INFO|MainThread|machine.py:145 - run()|File > "/usr/lib/python2.6/site-packages/resource_management/core/shell.py", line > 303, in _call > 2016-12-16 13:47:10,423|INFO|MainThread|machine.py:145 - run()|raise > ExecutionFailed(err_msg, code, out, err) > 2016-12-16 13:47:10,423|INFO|MainThread|machine.py:145 - > run()|ExecutionFailed: Execution of '/usr/bin/yum -d 0 -e 0 check > dependencies' returned 1. error: rpmdb: BDB0113 Thread/process > 16016/139791567193920 failed: BDB1507 Thread died in Berkeley DB library > 2016-12-16 13:47:10,424|INFO|MainThread|machine.py:145 - run()|error: db5 > error(-30973) from dbenv->failchk: BDB0087 DB_RUNRECOVERY: Fatal error, run > database recovery > 2016-12-16 13:47:10,424|INFO|MainThread|machine.py:145 - run()|error: cannot > open Packages index using db5 - (-30973) > 2016-12-16 13:47:10,424|INFO|MainThread|machine.py:145 - run()|error: cannot > open Packages database in /var/lib/rpm > 2016-12-16 13:47:10,424|INFO|MainThread|machine.py:145 - > run()|CRITICAL:yum.main: > 2016-12-16 13:47:10,424|INFO|MainThread|machine.py:145 - run()| > 2016-12-16 13:47:10,425|INFO|MainThread|machine.py:145 - run()|Error: rpmdb > open failed > 2016-12-16 13:47:10,425|INFO|MainThread|machine.py:145 - run()|Traceback > (most recent call last): > 2016-12-16 13:47:10,425|INFO|MainThread|machine.py:145 - run()|File > "/var/lib/ambari-agent/cache/custom_actions/scripts/install_packages.py", > line 469, in <module> > 2016-12-16 13:47:10,425|INFO|MainThread|machine.py:145 - > run()|InstallPackages().execute() > 2016-12-16 13:47:10,425|INFO|MainThread|machine.py:145 - run()|File > "/usr/lib/python2.6/site-packages/resource_management/libraries/script/script.py", > line 287, in execute > 2016-12-16 13:47:10,426|INFO|MainThread|machine.py:145 - run()|method(env) > 2016-12-16 13:47:10,426|INFO|MainThread|machine.py:145 - run()|File > "/var/lib/ambari-agent/cache/custom_actions/scripts/install_packages.py", > line 179, in actionexecute > 2016-12-16 13:47:10,426|INFO|MainThread|machine.py:145 - run()|raise > Fail("Failed to distribute repositories/install packages") > {code} > > > Diffs > ----- > > ambari-common/src/main/python/ambari_commons/process_utils.py PRE-CREATION > ambari-common/src/main/python/ambari_commons/shell.py 8d26599 > > Diff: https://reviews.apache.org/r/56059/diff/ > > > Testing > ------- > > test on live cluster. Still there are things to check > > > Thanks, > > Dmitro Lisnichenko > >
