Re: Lock expired?

[Mark Burgess]

There is a lock for the daemon and a lock per started cfagent process.
Not sure that version you are using, but you should probably upgrade
to 2.1.16.
M

On Tue, 2005-10-04 at 15:09 +0200, Aleksandar Ivanisevic wrote:
> Ok, so how do I tell what holds the second lock and why?
> 
> # cfshow --active 
> lock..xxx.cfexecd.cfexecd_1243 = Wed Jun 29 16:04:38 2005
> 
> lock.cfagent_conf.xxx.link.link_4052 = Tue Oct  4 13:07:30 2005
> 
> lock.db.localhost.cfenvd.cfenvd_1258 = Wed Jun 29 16:04:31 2005
> 
> # ps -ef | grep cf
> root      1767     1  0 Jun29 ?        00:00:03 cfservd
> root      1787     1  0 Jun29 ?        00:08:22 cfenvd
> root      1796     1  0 Jun29 ?        00:00:55 cfexecd
> root     28428 28296  0 14:24 pts/0    00:00:00 grep cf
> 
> In the cfengine.log around that time I see normal hourly run of cfagent
> and bunch of locks removed, but not that one.
> 
> In the log I also see:
> 
> # grep -v "normally" /var/cfengine/cfengine.runlog
> Mon Oct  3 11:07:28 2005:Lock expired, process
> killed:pid=18050:link:link
> Mon Oct  3 16:07:41 2005:Lock expired, process
> killed:pid=30495:link:link
> Mon Oct  3 18:07:47 2005:Lock expired, process killed:pid=3485:link:link
> Mon Oct  3 22:07:58 2005:Lock expired, process
> killed:pid=14332:link:link
> Tue Oct  4 00:08:02 2005:Lock expired, process
> killed:pid=19436:link:link
> Tue Oct  4 03:08:10 2005:Lock expired, process
> killed:pid=26301:link:link
> Tue Oct  4 07:07:21 2005:Lock expired, process killed:pid=2632:link:link
> Tue Oct  4 10:07:30 2005:Lock expired, process killed:pid=8651:link:link
> Tue Oct  4 12:03:41 2005:Lock expired, process
> killed:pid=14359:link:link
> 
> On Tue, 2005-10-04 at 13:52 +0200, Mark Burgess wrote:
> > Use cfshow to see the locks.
> > 
> > On Tue, 2005-10-04 at 12:28 +0200, Aleksandar Ivanisevic wrote:
> > > What does this mean?
> > > 
> > > Oct  4 10:27:21 eul0300305 cfengine:EUL0300305[26010]: Lock
> > > lock.cfagent_conf.EUL0300305.link.link_4052 expired...(after 119/60
> > > minutes)  
> > > 
> > > if I run cfrun without -A, I see some of the actions not being executed,
> > > I presume it is because of these locks. How can I see which locks are
> > > currently in place and who is holding them? if I do
> > > db_dump /var/cfengine/cfengine_lock_db I see only gibberish ;)
> > > 
> > > Hmm... I've read the thread "Re: shellcommands not executing - lock
> > > problem?" from February but I can't see where it could apply to me.
> > > 
> > > Any hints for troubleshooting this? db is db4 (Redhat EL3)
> > > 
> > > [root@EUL0300261 root2]# file /var/cfengine/cfengine_lock_db
> > > /var/cfengine/cfengine_lock_db: Berkeley DB (Btree, version 9, native
> > > byte-order)
> > >