---
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/55480/#review161455
---
Will create a JIRA for this as soon as it's back online.
- Aaron Wood
On Jan. 12, 2017, 11:23 p.m., Aaron Wood wrote:
>
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/55480/
> ---
>
> (Updated Jan. 12, 2017, 11:23 p.m.)
>
>
> Review request for mesos and Anand Mazumdar.
>
>
> Repository: mesos
>
>
> Description
> ---
>
> This fixes the segfault that occurs when an executor sets a UUID that's not a
> valid v4 UUID and sends it off to the agent:
>
> ```
> ABORT: (../../3rdparty/stout/include/stout/try.hpp:77): Try::get() but state
> == ERROR: Not a valid UUID
> *** Aborted at 1484262968 (unix time) try "date -d @1484262968" if you are
> using GNU date ***
> PC: @ 0x7efeb6101428 (unknown)
> *** SIGABRT (@0x36b7) received by PID 14007 (TID 0x7efeabd29700) from PID
> 14007; stack trace: ***
> @ 0x7efeb64a6390 (unknown)
> @ 0x7efeb6101428 (unknown)
> @ 0x7efeb610302a (unknown)
> @ 0x560df739fa6e _Abort()
> @ 0x560df739fa9c _Abort()
> @ 0x7efebb53a5ad Try<>::get()
> @ 0x7efebb5363d6 Try<>::get()
> @ 0x7efebbd84809
> mesos::internal::slave::validation::executor::call::validate()
> @ 0x7efebbb59b36 mesos::internal::slave::Slave::Http::executor()
> @ 0x7efebbc773b8
> _ZZN5mesos8internal5slave5Slave10initializeEvENKUlRKN7process4http7RequestEE1_clES7_
> @ 0x7efebbcb5808
> _ZNSt17_Function_handlerIFN7process6FutureINS0_4http8ResponseEEERKNS2_7RequestEEZN5mesos8internal5slave5Slave10initializeEvEUlS7_E1_E9_M_invokeERKSt9_Any_dataS7_
> @ 0x7efebbfb2aea std::function<>::operator()()
> @ 0x7efebcb158b8
> _ZZZN7process11ProcessBase6_visitERKNS0_12HttpEndpointERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEERKNS_5OwnedINS_4http7RequestNKUlRK6OptionINSD_14authentication20AuthenticationResultEEE0_clESN_ENKUlbE1_clEb
> @ 0x7efebcb1a10a
> _ZZZNK7process9_DeferredIZZNS_11ProcessBase6_visitERKNS1_12HttpEndpointERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEERKNS_5OwnedINS_4http7RequestNKUlRK6OptionINSE_14authentication20AuthenticationResultEEE0_clESO_EUlbE1_EcvSt8functionIFT_T0_EEINS_6FutureINSE_8ResponseEEERKbEEvENKUlS12_E_clES12_ENKUlvE_clEv
> @ 0x7efebcb1c5f8
> _ZNSt17_Function_handlerIFN7process6FutureINS0_4http8ResponseEEEvEZZNKS0_9_DeferredIZZNS0_11ProcessBase6_visitERKNS7_12HttpEndpointERKNSt7__cxx1112basic_stringIcSt11char_traitsIcESaIcEEERKNS0_5OwnedINS2_7RequestNKUlRK6OptionINS2_14authentication20AuthenticationResultEEE0_clEST_EUlbE1_EcvSt8functionIFT_T0_EEIS4_RKbEEvENKUlS14_E_clES14_EUlvE_E9_M_invokeERKSt9_Any_data
> @ 0x7efebb5ce8ca std::function<>::operator()()
> @ 0x7efebb5c4b27
> _ZZN7process8internal8DispatchINS_6FutureINS_4http8ResponseclIRSt8functionIFS5_vS5_RKNS_4UPIDEOT_ENKUlPNS_11ProcessBaseEE_clESI_
> @ 0x7efebb5d4e1e
> _ZNSt17_Function_handlerIFvPN7process11ProcessBaseEEZNS0_8internal8DispatchINS0_6FutureINS0_4http8ResponseclIRSt8functionIFS9_vS9_RKNS0_4UPIDEOT_EUlS2_E_E9_M_invokeERKSt9_Any_dataOS2_
> @ 0x7efebcb30baf std::function<>::operator()()
> @ 0x7efebcb13fd6 process::ProcessBase::visit()
> @ 0x7efebcb1f3c8 process::DispatchEvent::visit()
> @ 0x7efebb3ab2ea process::ProcessBase::serve()
> @ 0x7efebcb0fe8a process::ProcessManager::resume()
> @ 0x7efebcb0c5a3
> _ZZN7process14ProcessManager12init_threadsEvENKUt_clEv
> @ 0x7efebcb1ea34
> _ZNSt12_Bind_simpleIFZN7process14ProcessManager12init_threadsEvEUt_vEE9_M_invokeIJEEEvSt12_Index_tupleIJXspT_EEE
> @ 0x7efebcb1e98a
> _ZNSt12_Bind_simpleIFZN7process14ProcessManager12init_threadsEvEUt_vEEclEv
> @ 0x7efebcb1e91a
> _ZNSt6thread5_ImplISt12_Bind_simpleIFZN7process14ProcessManager12init_threadsEvEUt_vEEE6_M_runEv
> @ 0x7efeb6980c80 (unknown)
> @ 0x7efeb649c6ba start_thread
> @ 0x7efeb61d282d (unknown)
> Aborted (core dumped)
> ```
>
>
> Diffs
> -
>
> src/slave/validation.cpp abd9b1248
>
> Diff: https://reviews.apache.org/r/55480/diff/
>
>
> Testing
> ---
>
> Verified by making sure no segfault occurs when rebuilding Mesos with this
> fix and pointing our framework at it. Our framework is currently not
> generating v4 UUIDs which was causing the issue.
>
>
> Thanks,
>
> Aaron Wood
>
>