> On Jan. 15, 2020, 2:55 p.m., Andrei Sekretenko wrote:
> > src/slave/slave.cpp
> > Line 6583 (original), 6583 (patched)
> > <https://reviews.apache.org/r/72003/diff/1/?file=2207139#file2207139line6583>
> >
> >     If we are not validating that `ExecutorInfo` of a task does not have 
> > `generated_by_agent` set, maybe we can at least forcibly clear it? either 
> > here or in `Slave::runTask()`, and document (in the previous patch) that 
> > agent **ignores** this field when received from outside?
> 
> Benjamin Bannier wrote:
>     That would be possible, but I am unsure it would be worth it. We 
> currently do perform mutations on user-provided executors, the documentation 
> of the added field clearly indicates that it should not be set nor be relied 
> on by frameworks, and even if it was set it would only lead to potentially 
> misbehaving tasks without wider effect.
>     
>     I feel we shouldn't do this, but please reopen if you believe it is 
> critical for some reason.

Added rejection of framework-specified `generated_by_agent` in the previous 
patch.


- Benjamin


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/72003/#review219270
-----------------------------------------------------------


On Jan. 15, 2020, 4:12 p.m., Benjamin Bannier wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/72003/
> -----------------------------------------------------------
> 
> (Updated Jan. 15, 2020, 4:12 p.m.)
> 
> 
> Review request for mesos, Andrei Sekretenko and Benjamin Mahler.
> 
> 
> Bugs: MESOS-10084
>     https://issues.apache.org/jira/browse/MESOS-10084
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> This patch modifies the agent code so it uses
> `ExecutorInfo::generated_by_agent` instead of the previous heuristic to
> detect agent-generated executors for command tasks. For that we add
> recovery logic for upgrade scenarios to inject a correct value for
> recovered executors, and set the appropriate value for newly created
> executors.
> 
> 
> Diffs
> -----
> 
>   src/slave/slave.hpp 3d191dcbdd58e851ad5e86fbb01a890407409c64 
>   src/slave/slave.cpp 0005971717f5b90da368b45caad8e209ada95fa5 
> 
> 
> Diff: https://reviews.apache.org/r/72003/diff/2/
> 
> 
> Testing
> -------
> 
> `make check`
> 
> 
> Thanks,
> 
> Benjamin Bannier
> 
>

Reply via email to