Bug 1130856

Summary:	ifup can be invoked before the interface is up
Product:	Red Hat Enterprise Linux 6	Reporter:	K. Y. Srinivasan <kys>
Component:	initscripts	Assignee:	Lukáš Nykrýn <lnykryn>
Status:	CLOSED WORKSFORME	QA Contact:	qe-baseos-daemons
Severity:	high	Docs Contact:
Priority:	unspecified
Version:	6.6	CC:	abgupta, haiyangz
Target Milestone:	rc
Target Release:	---
Hardware:	x86_64
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2015-01-12 13:13:52 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:
Embargoed:

Description K. Y. Srinivasan 2014-08-18 03:54:42 UTC

Description of problem: On Azure we have situations where ifup is invoked  
prior to interface coming online. Ifup fails and this results in a VM with no networking.

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


How reproducible:


Steps to Reproduce:
1.
2.
3.

Actual results:


Expected results:


Additional info:

Comment 2 Lukáš Nykrýn 2014-08-18 08:17:09 UTC

Maybe we could backport https://git.fedorahosted.org/cgit/initscripts.git/commit/?id=832e3236a72f88ad85b78ea35ec3beb80313dd5d

Comment 3 Sitsofe Wheeler 2014-09-30 03:39:49 UTC

Thanks for CC'ing me KY.

I think I originally reported this over in bug #1095387 back in May... As I mentioned in the upstream of that (https://bugzilla.kernel.org/show_bug.cgi?id=78771 ) this might be resolved by the upstream kernel patch https://lkml.org/lkml/2014/8/28/649 .

Comment 4 Sitsofe Wheeler 2014-09-30 03:49:34 UTC

(I should also note the above only fights the cause in one case which was a regression. If you start a VM when the network really isn't available the previous comment will do nothing for you)

Comment 5 Sitsofe Wheeler 2014-09-30 04:20:57 UTC

Something weird is happening.

With a non-debug 3.17rc kernel after 14 successful reboots the following happened:

Sep 30 13:35:04 a systemd[1]: Reached target Login Prompts.
Sep 30 13:35:04 a dbus-daemon[416]: dbus[416]: [system] Successfully activated service 'org.freedesktop.systemd1'
Sep 30 13:35:04 a dbus[416]: [system] Successfully activated service 'org.freedesktop.systemd1'
Sep 30 13:35:04 a systemd-logind[415]: New seat seat0.
Sep 30 13:35:04 a systemd-logind[415]: Watching system buttons on /dev/input/event0 (Power Button)
Sep 30 13:35:04 a systemd-logind[415]: Watching system buttons on /dev/input/event2 (AT Translated Set 2 keyboard)
Sep 30 13:35:04 a systemd[1]: Started Login Service.
Sep 30 13:35:04 a network[419]: Bringing up loopback interface:  [  OK  ]
Sep 30 13:35:04 a network[419]: Bringing up interface eth0:
Sep 30 13:35:04 a dhclient[518]: DHCPDISCOVER on eth0 to 255.255.255.255 port 67 interval 7 (xid=0x5b778314)
Sep 30 03:49:01 a systemd[1]: Time has been changed
Sep 30 03:49:02 a kernel: psmouse serio1: trackpoint: IBM TrackPoint firmware: 0x01, buttons: 0/0
Sep 30 03:49:02 a kernel: input: TPPS/2 IBM TrackPoint as /devices/platform/i8042/serio1/input/input4
Sep 30 03:49:02 a dhclient[518]: DHCPOFFER from 10.1.0.2
Sep 30 03:49:06 a systemd[1]: Time has been changed
Sep 30 03:49:11 a systemd[1]: Time has been changed
Sep 30 03:49:16 a systemd[1]: Time has been changed
Sep 30 03:49:21 a systemd[1]: Time has been changed
Sep 30 03:49:26 a systemd[1]: Time has been changed
Sep 30 03:49:28 a chronyd[412]: Backward time jump detected! (correction 35162.9 +- 1.0 seconds)
Sep 30 03:49:31 a systemd[1]: Time has been changed
[...]
Sep 30 03:49:56 a systemd[1]: Time has been changed
Sep 30 03:50:01 a systemd[1]: Starting user-0.slice.
Sep 30 03:50:01 a systemd[1]: Created slice user-0.slice.
Sep 30 03:50:01 a systemd[1]: Starting User Manager for 0...
Sep 30 03:50:01 a systemd[1]: Starting Session 1 of user root.
Sep 30 03:50:01 a systemd[1]: Started Session 1 of user root.
Sep 30 03:50:01 a systemd[623]: pam_unix(systemd-user:session): session opened for user root by (uid=0)
Sep 30 03:50:01 a CROND[624]: (root) CMD (/usr/lib64/sa/sa1 1 1)
Sep 30 03:50:01 a systemd[623]: Failed to open private bus connection: Failed to connect to socket /run/user/0/dbus/user_bus_socket: No such file or directory
Sep 30 03:50:01 a systemd[623]: Starting Default.
Sep 30 03:50:01 a systemd[623]: Reached target Default.
Sep 30 03:50:01 a systemd[623]: Startup finished in 14ms.
Sep 30 03:50:01 a systemd[1]: Started User Manager for 0.
Sep 30 03:50:01 a systemd[623]: Time has been changed
Sep 30 03:50:01 a systemd[1]: Time has been changed
Sep 30 03:50:06 a systemd[1]: Time has been changed
Sep 30 03:50:06 a systemd[623]: Time has been changed
Sep 30 03:50:11 a systemd[623]: Time has been changed
[...]
Sep 30 03:53:01 a systemd[623]: Time has been changed
Sep 30 03:53:01 a systemd[1]: Time has been changed
Sep 30 03:53:06 a systemd[1]: Time has been changed
Sep 30 03:53:06 a systemd[623]: Time has been changed
Sep 30 03:54:00 a systemd[1]: network.service operation timed out. Terminating.
Sep 30 03:54:00 a systemd[1]: Failed to start LSB: Bring up/down networking.
Sep 30 03:54:00 a systemd[1]: Unit network.service entered failed state.
Sep 30 03:54:00 a systemd[1]: Starting Multi-User System.
Sep 30 03:54:00 a systemd[1]: Reached target Multi-User System.

So the DHCP server sent an offer but we fell on the floor anyway - why? Additionally, who is changing the time?

Comment 6 Lukáš Nykrýn 2014-09-30 07:25:25 UTC

Which version of systemd are you using?
We are not starting user instances of systemd in rhel7
Sep 30 03:50:01 a systemd[623]: Failed to open private bus connection: Failed to connect to socket /run/user/0/dbus/user_bus_socket: No such file or directory

Comment 7 Lukáš Nykrýn 2014-09-30 07:35:59 UTC

And this is a bug for rhel6, so please if you have other issues file a new bug for proper product and component.

Comment 8 Sitsofe Wheeler 2014-09-30 20:54:09 UTC

OK let me unsubscribe from this bug.