[
https://issues.apache.org/jira/browse/MESOS-6270?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15531360#comment-15531360
]
Jie Yu commented on MESOS-6270:
-------------------------------
I think the root cause is that we might call 'destroy' before parent/child
relationship is properly setup. This is because we call onAny(reaped) in
recover. Since recover is async, reaped might get invoked, leading to 'destroy'
being called.
> Agent crashes when trying to recover pods.
> ------------------------------------------
>
> Key: MESOS-6270
> URL: https://issues.apache.org/jira/browse/MESOS-6270
> Project: Mesos
> Issue Type: Bug
> Components: containerization
> Affects Versions: 1.1.0
> Reporter: Avinash Sridharan
> Assignee: Gilbert Song
> Labels: mesosphere
>
> The `MesosContainerizer` seems to be crashing when its trying to recover pods
> after a restart. Seems like the containerizer is unable to find the parent of
> a container it needs to destroy during recovery.
> ```
> vagrant@mesos-dev:~/mesosphere/mesos/build$ startagentcni
> WARNING: Logging before InitGoogleLogging() is written to STDERR
> I0928 23:46:49.638267 31124 main.cpp:243] Build: 2016-09-05 16:37:12 by
> vagrant
> I0928 23:46:49.639432 31124 main.cpp:244] Version: 1.1.0
> I0928 23:46:49.639801 31124 main.cpp:251] Git SHA:
> 97e48e4af0bf497cfe148bbeba2e2eace1a030d3
> I0928 23:46:49.642931 31124 process.cpp:1069] libprocess is initialized on
> 10.0.2.15:5051 with 8 worker threads
> I0928 23:46:49.645144 31124 logging.cpp:199] Logging to STDERR
> I0928 23:46:49.649516 31124 containerizer.cpp:226] Using isolation:
> filesystem/linux,docker/runtime,network/cni,volume/image
> I0928 23:46:49.654670 31124 linux_launcher.cpp:144] Using
> /sys/fs/cgroup/freezer as the freezer hierarchy for the Linux launcher
> E0928 23:46:49.662094 31124 shell.hpp:110] Command 'hadoop version 2>&1'
> failed; this is the output:
> sh: 1: hadoop: not found
> I0928 23:46:49.662240 31124 fetcher.cpp:69] Skipping URI fetcher plugin
> 'hadoop' as it could not be created: Failed to create HDFS client: Failed to
> execute 'hadoop version 2>&1'; the command was either not found or exited
> with a n
> on-zero exit status: 127
> I0928 23:46:49.662992 31124 registry_puller.cpp:111] Creating registry puller
> with docker registry 'https://registry-1.docker.io'
> I0928 23:46:49.676386 31142 slave.cpp:208] Mesos agent started on
> (1)@10.0.2.15:5051
> I0928 23:46:49.676834 31142 slave.cpp:209] Flags at startup:
> --appc_simple_discovery_uri_prefix="http://";
> --appc_store_dir="/tmp/mesos/store/appc" --authenticate_http_readonly="false"
> --authenticate_http_readwrite="false" --authent
> icatee="crammd5" --authentication_backoff_factor="1secs" --authorizer="local"
> --cgroups_cpu_enable_pids_and_tids_count="false" --cgroups_enable_cfs="false"
> --cgroups_hierarchy="/sys/fs/cgroup" --cgroups_limit_swap="false" --cgroups
> _root="mesos" --container_disk_watch_interval="15secs"
> --containerizers="mesos" --default_role="*" --disk_watch_interval="1mins"
> --docker="docker" --docker_kill_orphans="true"
> --docker_registry="https://registry-1.docker.io"; --dock
> er_remove_delay="6hrs" --docker_socket="/var/run/docker.sock"
> --docker_stop_timeout="0ns" --docker_store_dir="/tmp/mesos/store/docker"
> --docker_volume_checkpoint_dir="/var/run/mesos/isolators/docker/volume"
> --enforce_container_disk
> _quota="false" --executor_registration_timeout="1mins"
> --executor_shutdown_grace_period="5secs"
> --fetcher_cache_dir="/tmp/mesos/fetch" --fetcher_cache_size="2GB"
> --frameworks_home="" --gc_delay="1weeks" --gc_disk_headroom="0.1" --h
> adoop_home="" --help="false" --hostname_lookup="true"
> --http_authenticators="basic" --http_command_executor="false"
> --image_providers="docker" --image_provisioner_backend="copy"
> --initialize_driver_logging="true" --ip="10.0.2.15" -
> -isolation="filesystem/linux,docker/runtime" --launcher="linux"
> --launcher_dir="/home/vagrant/mesosphere/mesos/build/src" --logbufsecs="0"
> --logging_level="INFO" --master="10.0.2.15:5050"
> --network_cni_config_dir="/home/vagrant/cni
> /config" --network_cni_plugins_dir="/home/vagrant/dev/go/cni/bin"
> --oversubscribed_resources_interval="15secs" --perf_duration="10secs"
> --perf_interval="1mins" --port="5051" --qos_correction_interval_min="0ns"
> --quiet="false" --rec
> over="reconnect" --recovery_timeout="15mins"
> --registration_backoff_factor="1secs" --revocable_cpu_low_priority="true"
> --runtime_dir="/var/run/mesos" --sandbox_directory="/mnt/mesos/sandbox"
> --strict="true" --switch_user="true" --s
> ystemd_enable_support="true"
> --systemd_runtime_directory="/run/systemd/system" --version="false"
> --work_dir="/var/lib/mesos"
> I0928 23:46:49.679481 31142 slave.cpp:533] Agent resources: cpus(*):4;
> mem(*):6961; disk(*):35164; ports(*):[31000-32000]
> I0928 23:46:49.679832 31142 slave.cpp:541] Agent attributes: [ ]
> I0928 23:46:49.680160 31142 slave.cpp:546] Agent hostname: mesos-dev
> I0928 23:46:49.685725 31147 state.cpp:57] Recovering state from
> '/var/lib/mesos/meta'
> I0928 23:46:49.687319 31147 fetcher.cpp:86] Clearing fetcher cache
> I0928 23:46:49.687815 31147 status_update_manager.cpp:203] Recovering status
> update manager
> I0928 23:46:49.688199 31147 containerizer.cpp:580] Recovering containerizer
> I0928 23:46:49.692178 31143 linux_launcher.cpp:288] Recovered container
> 9e94e2b0-c07d-4641-9b65-f77b1601c696.7222dd27-576c-4f0e-afa3-17ae9f93539c
> I0928 23:46:49.692248 31143 linux_launcher.cpp:288] Recovered container
> 9e94e2b0-c07d-4641-9b65-f77b1601c696.39caa07f-6dd2-4acd-87be-d400467890d3
> I0928 23:46:49.692276 31143 linux_launcher.cpp:288] Recovered container
> 9e94e2b0-c07d-4641-9b65-f77b1601c696
> I0928 23:46:49.692312 31143 linux_launcher.cpp:370]
> 9e94e2b0-c07d-4641-9b65-f77b1601c696 is a known orphaned container
> I0928 23:46:49.692335 31143 linux_launcher.cpp:370]
> 9e94e2b0-c07d-4641-9b65-f77b1601c696.39caa07f-6dd2-4acd-87be-d400467890d3 is
> a known orphaned container
> I0928 23:46:49.692355 31143 linux_launcher.cpp:370]
> 9e94e2b0-c07d-4641-9b65-f77b1601c696.7222dd27-576c-4f0e-afa3-17ae9f93539c is
> a known orphaned container
> I0928 23:46:49.694280 31147 containerizer.cpp:2147] Container
> 9e94e2b0-c07d-4641-9b65-f77b1601c696 has exited
> I0928 23:46:49.694339 31147 containerizer.cpp:1835] Destroying container
> 9e94e2b0-c07d-4641-9b65-f77b1601c696
> I0928 23:46:49.695274 31145 linux_launcher.cpp:495] Asked to destroy
> container 9e94e2b0-c07d-4641-9b65-f77b1601c696
> I0928 23:46:49.696429 31146 metadata_manager.cpp:205] No images to load from
> disk. Docker provisioner image storage path
> '/tmp/mesos/store/docker/storedImages' does not exist
> I0928 23:46:49.697399 31147 provisioner.cpp:253] Provisioner recovery complete
> F0928 23:46:49.697675 31142 containerizer.cpp:867] Check failed:
> containers_.contains(containerId.parent())
> *** Check failure stack trace: ***
> F0928 23:46:49.697675 31142 containerizer.cpp:867] Check failed:
> containers_.contains(containerId.parent())
> *** Check failure stack trace: ***
> @ 0x7f2e3bdd514d google::LogMessage::Fail()
> @ 0x7f2e3bdd452e google::LogMessage::SendToLog()
> @ 0x7f2e3bdd4e0d google::LogMessage::Flush()
> @ 0x7f2e3bdd8288 google::LogMessageFatal::~LogMessageFatal()
> @ 0x7f2e3b0a7562
> mesos::internal::slave::MesosContainerizerProcess::__recover()
> @ 0x7f2e3b12dee8
> _ZZN7process8dispatchI7NothingN5mesos8internal5slave25MesosContainerizerProcessERKSt4listINS2_5slave14ContainerStateESaIS8_EERK7hashsetINS2_11ContainerIDESt4hashISE_ESt8equal_toISE_EESA_SJ_EENS_6FutureIT_EERKNS_3PIDIT0_EEMSQ_FSO_T1_T2_ET3_T4_ENKUlPNS_11ProcessBaseEE_clES11_
> @ 0x7f2e3b12da82
> _ZNSt17_Function_handlerIFvPN7process11ProcessBaseEEZNS0_8dispatchI7NothingN5mesos8internal5slave25MesosContainerizerProcessERKSt4listINS6_5slave14ContainerStateESaISC_EERK7hashsetINS6_11ContainerIDESt4hashISI_ESt8equal_toISI_EESE_SN_EENS0_6FutureIT_EERKNS0_3PIDIT0_EEMSU_FSS_T1_T2_ET3_T4_EUlS2_E_E9_M_invokeERKSt9_Any_dataS2_
> @ 0x7f2e3bd4cc68 std::function<>::operator()()
> @ 0x7f2e3bd34054 process::ProcessBase::visit()
> @ 0x7f2e3bd90b2e process::DispatchEvent::visit()
> @ 0x7f2e3a4f34c1 process::ProcessBase::serve()
> @ 0x7f2e3bd31d54 process::ProcessManager::resume()
> @ 0x7f2e3bd3cd0c
> process::ProcessManager::init_threads()::$_1::operator()()
> @ 0x7f2e3bd3cc15
> _ZNSt12_Bind_simpleIFZN7process14ProcessManager12init_threadsEvE3$_1vEE9_M_invokeIJEEEvSt12_Index_tupleIJXspT_EEE
> @ 0x7f2e3bd3cbe5 std::_Bind_simple<>::operator()()
> @ 0x7f2e3bd3cbbc std::thread::_Impl<>::_M_run()
> @ 0x7f2e3765ca60 (unknown)
> @ 0x7f2e3717f182 start_thread
> @ 0x7f2e36eac47d (unknown)
> ```
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
