mesos-reviews mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Benjamin Mahler <bmah...@apache.org>
Subject Review Request 62049: Fixed an OOM due to a send loop for SSL sockets.
Date Sat, 02 Sep 2017 00:38:03 GMT

-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/62049/
-----------------------------------------------------------

Review request for mesos, Benno Evers, Chun-Hung Hsiao, Gilbert Song, Greg Mann, Jie Yu, and
Joseph Wu.


Bugs: MESOS-7934
    https://issues.apache.org/jira/browse/MESOS-7934


Repository: mesos


Description
-------

Per MESOS-7934, the LibeventSSLSocket incorrectly returns 0 to the
sender when an EOF or "dirty" SSL shutdown occurs (i.e. TCP close
before SSL close). Not only is this incorrect due to the caller
re-sending the same data again. In some cases (one is reproduced
in a follow-up test), this can trigger a loop in libprocess that
rapidly leads to an OOM.

The fix here is to fail the send instead. This is the best we can
do with libevent 2.0.x since it does not set BEV_EVENT_READING or
BEV_EVENT_WRITING for SSL buffevents. With libevent 2.1.x, we can
fix our logic to deal with the read and write side events separately.

https://github.com/libevent/libevent/commit/f7eb69ace

Comments are added in a follow up change to explain this for
posterity, and MESOS-7930 tracks the additional tech debt that
needs to be addressed for SSL socket support.


Diffs
-----

  3rdparty/libprocess/src/libevent_ssl_socket.cpp 0fa7565d1dc49e53890708df154c4d8ea278ad37



Diff: https://reviews.apache.org/r/62049/diff/1/


Testing
-------

Tested in subsequent patch.


Thanks,

Benjamin Mahler


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message