You are here
EGY - Wed, 2014/03/12 - 13:07
Hello,
granted that I'm writing about a ubuntu-10.04-turnkey-lamp_11.3-1_i386 template, virtualized and contained in a CT of a Proxmox Host, I'm facing in a strange issue after a monthly-scheduled upgrade of this distro happened two months ago (December 2013).
First of all, I've to say that I've token notice of this issue, which I'm about to expose, just after a power-failure happened 3 weeks ago. Then, networking service wasn't able to get start up at the system boot, whereas it was able to do it manually.
After a lof of "googling" and comparing this distro upgrated with an identical one but not upgrated, I've understood that it depends on the upgrade of two booting managements present in my distro,SysVInit and UpStart, which leads them in conflict with each other.
By an empirical method, I tried to bring init.d directory back at its original state recreating symbolic links to the upstart-job. So, it appears as follow:
root@VMLAMP01 /etc/init.d# ls -l total 368 -rw-r--r-- 1 root root 2427 Oct 15 2012 README lrwxrwxrwx 1 root root 21 Mar 8 16:02 acpid -> /lib/init/upstart-job -rw-r--r-- 1 root root 2227 Apr 15 2013 acpid.orig -rwxr-xr-x 1 root root 7820 Jan 31 19:35 apache2 -rwxr-xr-x 1 root root 2341 Sep 7 2009 bootlogd -rwxr-xr-x 1 root root 1276 Oct 15 2012 bootlogs -rwxr-xr-x 1 root root 1281 Jul 14 2013 bootmisc.sh -rwxr-xr-x 1 root root 4027 Jun 28 2010 casper -rwxr-xr-x 1 root root 3816 Jul 14 2013 checkfs.sh -rwxr-xr-x 1 root root 1099 Jul 14 2013 checkroot-bootclean.sh -rwxr-xr-x 1 root root 9673 Jul 14 2013 checkroot.sh -rwxr-xr-x 1 root root 1780 Mar 29 2013 confconsole lrwxrwxrwx 1 root root 21 Mar 8 16:11 console-setup -> /lib/init/upstart-job -rw-r--r-- 1 root root 1379 Dec 9 2011 console-setup.orig lrwxrwxrwx 1 root root 21 Mar 8 16:03 cron -> /lib/init/upstart-job -rw-r--r-- 1 root root 3033 Jul 3 2012 cron.orig -rwxr-xr-x 1 root root 2182 Mar 29 2013 di-live lrwxrwxrwx 1 root root 21 Mar 8 16:11 dmesg -> /lib/init/upstart-job -rwxr-xr-x 1 root root 1105 Nov 23 2010 grub-common -rwxr-xr-x 1 root root 1329 Sep 7 2009 halt -rwxr-xr-x 1 root root 10577 Jun 24 2012 hdparm lrwxrwxrwx 1 root root 21 Mar 8 16:12 hostname -> /lib/init/upstart-job -rw-r--r-- 1 root root 1423 Oct 15 2012 hostname.sh.orig -rwxr-xr-x 1 root root 1359 Jun 21 2011 hubdns lrwxrwxrwx 1 root root 21 Mar 8 16:14 hwclock -> /lib/init/upstart-job lrwxrwxrwx 1 root root 21 Mar 8 16:14 hwclock-save -> /lib/init/upstart-job -rwxr-xr-x 1 root root 3882 Feb 24 12:52 hwclock.sh.orig -rwxr-xr-x 1 root root 1164 Nov 1 05:11 inithooks -rwxr-xr-x 1 root root 1019 Feb 25 14:27 jboss5 -rwxr-xr-x 1 root root 12550 Sep 21 2012 jetty -rwxr-xr-x 1 root root 12357 Apr 25 2012 jetty.dpkg-dist -rwxr-xr-x 1 root root 7592 Apr 28 2012 kbd -rwxr-xr-x 1 root root 1591 Oct 1 2012 keyboard-setup -rwxr-xr-x 1 root root 1293 Sep 7 2009 killprocs lrwxrwxrwx 1 root root 21 Aug 13 2012 module-init-tools -> /lib/init/upstart-job -rwxr-xr-x 1 root root 995 Oct 15 2012 motd -rwxr-xr-x 1 root root 670 Feb 24 2013 mountall-bootclean.sh -rwxr-xr-x 1 root root 2128 Feb 24 2013 mountall.sh -rwxr-xr-x 1 root root 1508 Jul 14 2013 mountdevsubfs.sh -rwxr-xr-x 1 root root 1414 Feb 24 15:07 mountkernfs.sh -rwxr-xr-x 1 root root 678 Feb 24 2013 mountnfs-bootclean.sh -rwxr-xr-x 1 root root 2440 Oct 15 2012 mountnfs.sh -rwxr-xr-x 1 root root 1731 Jul 14 2013 mtab.sh -rw-r--r-- 1 root root 5437 Dec 13 18:24 mysql.orig lrwxrwxrwx 1 root root 21 Mar 8 16:16 network-interface -> /lib/init/upstart-job lrwxrwxrwx 1 root root 21 Mar 8 16:16 network-interface-security -> /lib/init/upstart-job -rw-r--r-- 1 root root 4323 Feb 25 18:45 networking -rwxr-xr-x 1 root root 1818 Feb 3 2010 ntp -rwxr-xr-x 1 root root 882 Sep 7 2009 ondemand lrwxrwxrwx 1 root root 21 Mar 8 16:17 plymouth -> /lib/init/upstart-job lrwxrwxrwx 1 root root 21 Mar 8 16:18 plymouth-log -> /lib/init/upstart-job -rw-r--r-- 1 root root 725 Jun 30 2012 plymouth-log.orig lrwxrwxrwx 1 root root 21 Mar 8 16:18 plymouth-splash -> /lib/init/upstart-job lrwxrwxrwx 1 root root 21 Mar 8 16:18 plymouth-stop -> /lib/init/upstart-job -rw-r--r-- 1 root root 1313 Jun 30 2012 plymouth.orig -rwxr-xr-x 1 root root 7358 Mar 11 2013 postfix lrwxrwxrwx 1 root root 21 Aug 13 2012 procps -> /lib/init/upstart-job -rwxr-xr-x 1 root root 3093 Jan 18 2010 quota -rwxr-xr-x 1 root root 1823 Jan 18 2010 quotarpc -rwxr-xr-x 1 root root 6120 Oct 15 2012 rc -rwxr-xr-x 1 root root 782 Oct 15 2012 rc.local -rwxr-xr-x 1 root root 117 Oct 15 2012 rcS -rwxr-xr-x 1 root root 639 Sep 7 2009 reboot -rwxr-xr-x 1 root root 1074 Jul 14 2013 rmnologin -rwxr-xr-x 1 root root 4395 Dec 2 2012 rsync lrwxrwxrwx 1 root root 21 Mar 8 16:19 rsyslog -> /lib/init/upstart-job -rw-r--r-- 1 root root 3054 Mar 5 2013 rsyslog.orig -rwxr-xr-x 1 root root 1163 Sep 16 2012 screen-cleanup -rwxr-xr-x 1 root root 3200 Oct 15 2012 sendsigs -rwxr-xr-x 1 root root 4770 Apr 2 2013 shellinabox -rwxr-xr-x 1 root root 590 Sep 7 2009 single -rw-r--r-- 1 root root 4290 Oct 15 2012 skeleton -rwxr-xr-x 1 root root 3881 Feb 8 2013 ssh -rwxr-xr-x 1 root root 519 Sep 7 2009 stop-bootlogd -rwxr-xr-x 1 root root 1095 Sep 7 2009 stop-bootlogd-single -rwxr-xr-x 1 root root 714 Jun 28 2012 sudo -rwxr-xr-x 1 root root 4572 Nov 1 05:11 turnkey-init-fence lrwxrwxrwx 1 root root 21 Mar 8 16:19 udev -> /lib/init/upstart-job lrwxrwxrwx 1 root root 21 Mar 8 16:20 udev-finish -> /lib/init/upstart-job -rwxr-xr-x 1 root root 1179 Aug 19 2012 udev-mtab -rw-r--r-- 1 root root 8827 Mar 7 17:01 udev.orig lrwxrwxrwx 1 root root 21 Mar 8 16:20 udevmonitor -> /lib/init/upstart-job lrwxrwxrwx 1 root root 21 Mar 8 16:20 udevtrigger -> /lib/init/upstart-job -rwxr-xr-x 1 root root 2721 Apr 10 2013 umountfs -rwxr-xr-x 1 root root 2195 Apr 10 2013 umountnfs.sh -rwxr-xr-x 1 root root 1122 Oct 15 2012 umountroot lrwxrwxrwx 1 root root 21 Mar 8 16:17 upstart-job -> /lib/init/upstart-job -rwxr-xr-x 1 root root 3111 Oct 15 2012 urandom -rwxr-xr-x 1 root root 1359 Jul 28 2013 webmin -rwxr-xr-x 1 root root 1342 Apr 11 2012 winbind -rwxr-xr-x 1 root root 2666 Mar 2 2012 x11-common
where *.orig are files I've replaced by a sym link.
Now, I achieve networking to start up at system boot. But, there is another strange behavior of mysql service: sometime it start up at boot system but sometime it doesn't, and I don't know why!!!
In fine, after this distro upgrade, I have issues with python too. Here is an example:
root@VMLAMP01 /etc/init.d# turnkey-sysinfo
Traceback (most recent call last):
File "/usr/bin/turnkey-sysinfo", line 19, in <module>
import netinfo
ImportError: No module named netinfo
I searched this module, netinfo, and I found:
root@VMLAMP01 /etc/init.d# find /usr/lib/ -name "net*" /usr/lib/tklbam/deps/lib/python2.7/dist-packages/boto/ec2/networkinterface.py /usr/lib/jvm/java-6-openjdk/jre/lib/net.properties /usr/lib/python2.7/dist-packages/netinfo.pyc /usr/lib/python2.7/dist-packages/netinfo.py /usr/lib/python3.2/__pycache__/netrc.cpython-32.pyc /usr/lib/python3.2/netrc.py /usr/lib/python2.6/netrc.pyc /usr/lib/python2.6/netrc.py
But, my system uses 2.6 version of python:
root@VMLAMP01 /etc/init.d# which python /usr/bin/python root@VMLAMP01 /etc/init.d# ls -l /usr/bin/python lrwxrwxrwx 1 root root 9 Aug 13 2012 /usr/bin/python -> python2.6
thus far, I've exposed what are my issues... Now, I write what are my questions:
- how can I know why my mysql service sometime starts up and sometime not?
- how can I exactly know which services I have to start up by Up Start system and which by SysVInit?
- how can I fix python issues? (bearing in mind that not all modules of 2.6 version are in 2.7 version)
Please, help me.
Forum:
I can't give you specifics
But I recall that there was a bug with the MySQL upstart on the default 10.04. I vaguely recall that TKL implemented a workaround although I don't recall any details...
Sorry it's probably not a great help, but hopefully might be a useful hint...?!
Tracing UpStart and SysVInit
I think it doesn't depend on a bug of MySQL upstart on the default 10.04... because I created this CT in ProXMoX more than 1 year ago and I faced this issue just after last DISTRO upgrade: I've never had any problems before.
Anyway, could you suggest me how can I trace UpStart and SysInit processes at the boot of S.O.?
I have a idea, but I can test it because I can't able to get trace of UpStart jobs.
I read somewhere that was written a job named openvz.conf just to make UpStart service able to run correct.
root@VMLAMP01 /etc/init# cat openvz.conf # OpenVZ - Fix init sequence to have OpenVZ working with upstart # Source: http://blog.bodhizazen.net/linux/ubuntu-10-04-openvz-templates/ description "Fix OpenVZ" start on startup task pre-start script mount -t devpts devpts /dev/pts || true mount -t tmpfs varrun /var/run mount -t tmpfs varlock /var/lock mkdir -p /var/run/network if [ ! -e /etc/mtab ]; then cat /proc/mounts > /etc/mtab fi touch /var/run/utmp chmod 664 /var/run/utmp chown root.utmp /var/run/utmp if [ "$(find /etc/network/ -name upstart -type f)" ]; then chmod -x /etc/network/*/upstart || true fi end script script start networking initctl emit filesystem --no-wait initctl emit local-filesystems --no-wait initctl emit virtual-filesystems --no-wait initctl emit net-device-up --no-wait init 2 end scriptAnd here it is wrote
command! So, if we go in mysql.conf,
root@VMLAMP01 /etc/init# cat mysql.conf # MySQL Service description "MySQL Server" author "Mario Limonciello <superm1@ubuntu.com>" start on (net-device-up and local-filesystems and runlevel [2345]) stop on runlevel [016]we can see that it start when system goes in runlevel 2. Therefore, this is my idea:
could it be that upstart-job openvz is not able, sometime, to force the system to go in runlevel 2??
This can be a breakthrough, but I need to know how I can trace init process before.
I also read that it is possible debugging of Up-Start by adding "--verbose" at kernel, but it is an impossible modification in a OpenVZ Virtual Machine.
Could you help me, please?
Best Regards.
Sorry I'm not familar with upstart at all
I'm inclined to disagree that your experience rules out the possibility of the MySQL upstart bug reappearing due to your latest upgrade. My suspicion is/was that this upgrade (for whatever reason...) has undone whatever the MySQL bug workaround was and reverted to the original setup. Perhaps your previous upgrades didn't mess with the fix, but this one did...? But really it's academic because I'm only guessing...
I haven't had any contact with upstart since TKL v11 (which was a while ago now...) so realistically I'm not going to be much use to you. And even back then I probably wouldn't have been much help to you as I have learned as I have gone... Personally I would just migrate your content to a new v13.0 server and be done with it (IMO Ubuntu is a bit flakey; Debian is much more solid and stable)...
Otherwise, if you think that it is an Ubuntu issue then you're probably more likely to find someone who can meaningfully help you over on the Ubuntu forums (not meaning to fob you off, just being realistic).
Add new comment