Bug 1330853

Summary: Crash in docker-forward-journald
Product: [Fedora] Fedora Reporter: Nick Coghlan <ncoghlan>
Component: dockerAssignee: Lokesh Mandvekar <lsm5>
Status: CLOSED INSUFFICIENT_DATA QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: unspecified Docs Contact:
Priority: unspecified    
Version: 23CC: adimania, admiller, amurdaca, dwalsh, ichavero, jcajka, jchaloup, jhonce, lsm5, marianne, miminar, nalin, riek, vbatts
Target Milestone: ---   
Target Release: ---   
Hardware: Unspecified   
OS: Unspecified   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2016-04-27 08:15:58 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:

Description Nick Coghlan 2016-04-27 07:03:25 UTC 
Description of problem:

The Docker daemon crashed when running a "docker-compose up"

Version-Release number of selected component (if applicable):

docker-1.10.2-9.git0f5ac89.fc23.x86_64
docker-forward-journald-1.10.2-9.git0f5ac89.fc23.x86_64

The relevant part of the Go stack trace from journald appears to be:

goroutine 29152 [syscall, locked to thread]:
runtime.cgocall(0x10722b0, 0xc8238f9e60, 0xc800000000)
/usr/lib/golang/src/runtime/cgocall.go:120 +0x11b fp=0xc8238f9e30 sp=0xc8238f9e00
github.com/docker/docker/daemon/logger/journald._Cfunc_sd_journal_close(0x7f75f4003060)
github.com/docker/docker/daemon/logger/journald/_obj/_cgo_gotypes.go:153 +0x31 fp=0xc823
github.com/docker/docker/daemon/logger/journald.(*journald).readLogs(0xc822350ee0, 0xc82
/builddir/build/BUILD/docker-0f5ac89062e1f80bd87e7db9a94859adf7a188a7/_build/src/github.
runtime.goexit()
/usr/lib/golang/src/runtime/asm_amd64.s:1721 +0x1 fp=0xc8238f9f90 sp=0xc8238f9f88
created by github.com/docker/docker/daemon/logger/journald.(*journald).ReadLogs
/builddir/build/BUILD/docker-0f5ac89062e1f80bd87e7db9a94859adf7a188a7/_build/src/github.
Comment 1 Antonio Murdaca 2016-04-27 07:52:17 UTC 
The docker version you're using is pretty old. Could you update and see if you can reproduce? As we've fixed some stuff going on with the journald log driver in 1.10.3
Comment 2 Nick Coghlan 2016-04-27 08:15:58 UTC 
Sebastian van Stijn also pointed out https://bugzilla.redhat.com/show_bug.cgi?id=1300076, so I'm going to close this as insufficient data - I've only ever seen it once, and there have been some known issues with the docker/journald interface fixed in later versions.
Comment 3 Antonio Murdaca 2016-04-27 08:21:42 UTC 
That isn't the same issue AFAICT, but yes, pls upgrade and if you can reproduce just reopen this BZ :) Thx!
Comment 4 Daniel Walsh 2016-04-27 12:22:49 UTC 
Jon do you recognize this issue as something you fixed?
Comment 5 Nalin Dahyabhai 2016-04-28 16:13:33 UTC 
A crash in sd_journal_close() or sd_journal_next() during a cgo call is almost certainly the same as bug #1314463.