Hi, all Recently I’ve started using latest OpenNebula 5.0.1 2-node HA cluster (corosync + pacemaker) with the CEPH “Jewel” backend and everything works as expected, but I’ve met a strange issue with the instability of the “oned” process, when it randomly crashes on two completely different hardware…

It is indeed, can you give us more details? is there any relevant info in the logs? Maybe you have the core in the system (e.g. coredumpctl…)?

Hi, Ruben! Sorry, I didn’t find any dump file :frowning: And here we go with logs :slight_smile: oned.log: Tue Jul 19 09:03:07 2016 [Z0][MKP][D]: Monitoring marketplace OpenNebula Public (0) Tue Jul 19 09:03:09 2016 [Z0][MKP][D]: Marketplace OpenNebula Public (0) successfully monitored. Tue Ju…

I think you are being hit by this one: http://dev.opennebula.org/projects/opennebula/repository/revisions/63e736ab3de29f3db69d4f02e96985f115e5e961 This is now fixed in the repo. and will be released in a maintenance release this month

Do you think that this parsing error can lead to the crash of the entire “oned” process? Of course, I’ll try a new maintenance release immediately when it arrives, but can I somehow fix this error now? And another question - can I configure OpenNebula to write a dump when one of it’s processess wi…

same problem here. We try to have coredump too with no luck for the moment. So if any one has a suggestion that would be great. We configured the service to restart on abnormal-exit, will see if it works.

Hi, Edouard How did you configured the service to restart on abnormal exit? Do you have a Linux HA in the place or just a simple setup? -----Исходное сообщение----- От: “Edouard” opennebula@discoursemail.com Отправлено: ‎19.‎07.‎2016 20:27

Here is my /lib/systemd/system/opennebula.service : [Unit] Description=OpenNebula Cloud Controller Daemon After=syslog.target After=network.target After=remote-fs.target After=mariadb.service Before=opennebula-scheduler.service BindTo=opennebula-scheduler.service [Service] Type=simple Group=onead…

Thanks for sharing, I’ll try this config to see what will happen when oned will crash. It’s a pity that I can’t see or find any dependency with the segfault, and this happens randomly within 1 - 4 days. I can only see that both servers doesn’t have any load and this happens during not working hour…

Again have caught segfault, this time no errors about marketplace monitoring: messages log Jul 21 07:01:53 nebula-2 kernel: oned[15578]: segfault at 7fedf42c2000 ip 00007fee40d2411b sp 00007fee21ffa598 error 4 in libc-2.17.so [7fee40bda000+1b7000] Jul 21 07:01:53 nebula-2 systemd: opennebula.servi…

Hi, Same problem here… Opennebula 5.2.1 with random oned segfault crashes: [1207677.603276] oned[54334]: segfault at 7f4c68000000 ip 00007f4c6a9da982 sp 00007f4c6cf2c850 error 4 in libc-2.23.so[7f4c6a957000+1bf000] SO is Ubuntu 16.04, Kernel 4.4.0-78, Libc 2.23-ubuntu7 Any idea? Thanks!

Oned process segfault (error 4 in libc-2.17.so)

Product Support

TomK (TomK) March 25, 2018, 8:53pm 14

I’m having this problem as well via however I can reproduce it at will:

Topic		Replies	Views
OpenNebula Setup: One VM + 2 Physicals: Program terminated with signal 11, Segmentation fault Product Support	9	2355	November 26, 2019
Oned segmentation fault on Debian Jessie Product Support	3	1175	September 21, 2016
OpenNebula suddenly dies after upgrade to 5.4 Product Support	8	1112	August 2, 2017
OpenNebula 5.2 segfault Integration Support	3	825	January 2, 2017
Bootstrapping OpenNebula 5.4 from source Product Support	4	2666	August 1, 2017

Oned process segfault (error 4 in libc-2.17.so)

Related topics