[MESOS-6246] Libprocess links will not generate an ExitedEvent if the socket creation fails

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 0.27.3, 0.28.2, 1.0.1
• Fix Version/s: 0.28.3, 1.0.2, 1.1.0
• Component/s: libprocess
• Labels: libprocess mesosphere
• Story Points: 2
• Assignee: Joseph Wu
• Reporter: Joseph Wu
• Shepherd: Benjamin Mahler
• Created: 24/Sep/16 00:11
• Updated: 28/Sep/16 21:45
• Resolved: 28/Sep/16 21:45

Description

Noticed this while inspecting nearby code for potential races.

Normally, when a libprocess actor (the "linkee") links to a remote process, it does the following:
1) Create a socket.
2) Connect to the remote process (asynchronous).
3) Check the connection succeeded.

If (2) or (3) fail, the linkee will receive a ExitedEvent, which indicates that the link broke. In case (1) fails, there is no ExitedEvent:
https://github.com/apache/mesos/blob/7c833abbec9c9e4eb51d67f7a8e7a8d0870825f8/3rdparty/libprocess/src/process.cpp#L1558-L1562

Issue Links

• is related to: MESOS-6234 Potential socket leak during Zookeeper network changes