| Newsgroups |
gmane.comp.sysutils.cfengine.general |
| Message-ID |
<[email protected]> |
The "cf_lock.lmdb: Invalid argument" is a known issue that shouldn't have
too big of an impact. Mostly it occurs infrequently during bootstrap and no
other time.
In regards to nothing in /var/cfengine/outputs, do you get anything logged
if you do "cf-agent --no-lock"? cf-execd should be running cf-agent with
default logging level of notice and above (critical, error and warning) so
if a manual agent run with default logging doesn't output anything, nothing
would be in outputs. If nothing is output from the agent run then a file
will not be created.
A couple of things to debug the situation:
grep cf-execd /var/log/messages
or
systemctl stop cf-execd
/var/cfengine/bin/cf-execd --no-fork --verbose | less
/tmp/cf-execd-verbose.log
or adding a simple report to your policy like in services/main.cf simply add
reports:
"agent run at ${sys.date}";
As that should definitely be present in logs in /var/cfengine/outputs.
On Thursday, July 8, 2021 at 1:18:42 PM UTC-5 Beto wrote:
> The defunct cf-execd process goes away after a short period of time. It
> reappears after the cf-agent process ends. But now there is nothing logged
> in /var/cfengine/ouputs after multiple runs, although I know for a fact
> cf-agent did perform updates.
>
> On Thursday, July 8, 2021 at 12:56:03 PM UTC-5 Beto wrote:
>
>> Now I see this:
>>
>> root@baker1 [outputs]# r ps
>> ps -ef|grep c[f]-
>> root 7878 22661 0 12:50 ? 00:00:00 [cf-execd] <defunct>
>> root 22660 1 1 11:59 ? 00:00:44
>> /var/cfengine/bin/cf-monitord --no-fork
>> root 22661 1 0 11:59 ? 00:00:00
>> /var/cfengine/bin/cf-execd --no-fork
>> root 22662 1 0 11:59 ? 00:00:00
>> /var/cfengine/bin/cf-serverd --no-fork
>>
>> Which may be a result of this: (?)
>>
>> root@baker1 [outputs]# cat previous
>> error: Could not open database txn /var/cfengine/state/cf_lock.lmdb:
>> Invalid argument
>> On Thursday, July 8, 2021 at 12:04:42 PM UTC-5 Beto wrote:
>>
>>> I removed and reinstalled cfengine-community 3.18.0 and that seems to
>>> have fixed the cf-execd issue. Weird.
>>>
>>> On Thursday, July 8, 2021 at 9:22:30 AM UTC-5 Beto wrote:
>>>
>>>> root@baker1 [cfengine3.service.wants]# cat cf-execd.service
>>>> [Unit]
>>>> Description=CFEngine Enterprise Execution Scheduler
>>>> After=syslog.target
>>>> ConditionPathExists=@bindir@/cf-execd
>>>> ConditionPathExists=@workdir@/inputs/promises.cf
>>>> PartOf=cfengine3.service
>>>>
>>>> [Service]
>>>> Type=simple
>>>> ExecStart=@bindir@/cf-execd --no-fork
>>>> Restart=always
>>>> RestartSec=10
>>>> KillMode=process
>>>>
>>>> [Install]
>>>> WantedBy=multi-user.target
>>>> WantedBy=cfengine3.service
>>>>
>>>> On Thursday, July 8, 2021 at 9:18:45 AM UTC-5 [email protected]
>>>> wrote:
>>>>
>>>>> I checked an install of community on centos-7 and it would seem that
>>>>> the package is not corrupt but that somehow your cf-execd.service file did
>>>>> become corrupt.
>>>>>
>>>>> ```
>>>>> cf-remote spawn --platform centos-7-x64 --count 1 --role client --name
>>>>> cent7
>>>>> cf-remote --version 3.18.0 install --edition community --clients cent7
>>>>> ```
>>>>>
>>>>> I bootstrap and seem to have no issues.
>>>>>
>>>>> ```
>>>>> # grep cf-server /var/log/messages
>>>>> ...
>>>>> Jul 8 14:15:58 ip-172-31-18-118 cf-serverd: notice: Server is
>>>>> starting...
>>>>> Jul 8 14:15:58 ip-172-31-18-118 cf-serverd[11219]: CFEngine(server)
>>>>> Server is starting...
>>>>> ```
>>>>>
>>>>> Can you include the text of your cf-execd.service file?
>>>>>
>>>>> Thanks,
>>>>> Craig
>>>>> On Thursday, July 8, 2021 at 8:02:15 AM UTC-5 Beto wrote:
>>>>>
>>>>>> ut-oh, may have popped the cork a bit too quick. Now cf-execd fails
>>>>>> to start. Looking at cf-execd.service, it contains the text of a symlink:
>>>>>>
>>>>>> root@baker1 [cfengine3.service.wants]# cd /usr/lib/systemd/system/
>>>>>> root@baker1 [system]# ll cf*
>>>>>> -rw-r--r--. 1 root root 462 Jun 23 10:53 cf-apache.service
>>>>>> -rw-r--r--. 1 root root 864 Jun 23 10:53 cfengine3.service
>>>>>> -rw-r--r--. 1 root root 101 Jul 7 09:28 cf-execd.service
>>>>>> -rw-r--r--. 1 root root 443 Jun 23 10:53 cf-hub.service
>>>>>> -rw-r--r--. 1 root root 358 Jun 23 10:53 cf-monitord.service
>>>>>> -rw-r--r--. 1 root root 1150 Jun 23 10:53 cf-postgres.service
>>>>>> -rw-r--r--. 1 root root 664 Jun 23 10:53 cf-runalerts.service
>>>>>> -rw-r--r--. 1 root root 437 Jun 23 10:53 cf-serverd.service
>>>>>> root@baker1 [system]# cat cf-execd.service
>>>>>> lrwxrwxrwx. 1 root root 40 Dec 8 2017 cf-execd.service ->
>>>>>> /usr/lib/systemd/system/cf-execd.service
>>>>>>
>>>>>> So I replaced this with cf-execd.service from github but still not
>>>>>> working. This is from the journal:
>>>>>>
>>>>>> systemd: Cannot add dependency job for unit cf-execd.service,
>>>>>> ignoring: Unit is not loaded properly: Invalid argument.
>>>>>>
>>>>>> On Wednesday, July 7, 2021 at 1:40:37 PM UTC-5 Beto wrote:
>>>>>>
>>>>>>> Yes sir, ipv6 is disabled..
>>>>>>>
>>>>>>> On Wednesday, July 7, 2021 at 1:19:45 PM UTC-5
>>>>>>> [email protected] wrote:
>>>>>>>
>>>>>>>> Perfect. Thank you. The issue is with the bind interface of "::"
>>>>>>>> aka ipv6 wildcard address, if this fails then we don't fall back to ipv4.
>>>>>>>>
>>>>>>>> Do you by chance have ipv6 disabled? I would suspect the answer is
>>>>>>>> yes but would like to know for sure.
>>>>>>>>
>>>>>>>> Thanks,
>>>>>>>> Craig
>>>>>>>>
>>>>>>>> On Wednesday, July 7, 2021 at 1:05:56 PM UTC-5 Beto wrote:
>>>>>>>>
>>>>>>>>> Oh btw this is RHEL7.
>>>>>>>>>
>>>>>>>>> On Wednesday, July 7, 2021 at 1:00:32 PM UTC-5 Beto wrote:
>>>>>>>>>
>>>>>>>>>> That did it. Thanks, Craig!
>>>>>>>>>>
>>>>>>>>>> On Wednesday, July 7, 2021 at 12:45:10 PM UTC-5
>>>>>>>>>> [email protected] wrote:
>>>>>>>>>>
>>>>>>>>>>> Can you include some basic information like OS name/version?
>>>>>>>>>>> Thanks!
>>>>>>>>>>>
>>>>>>>>>>> Yes. This is likely a known issue we are working through right
>>>>>>>>>>> now.
>>>>>>>>>>>
>>>>>>>>>>> Please remove or comment out this section of policy and try
>>>>>>>>>>> again.
>>>>>>>>>>>
>>>>>>>>>>>
>>>>>>>>>>> https://github.com/cfengine/masterfiles/blob/3.18.x/controls/cf_serverd.cf#L73-L77
>>>>>>>>>>>
>>>>>>>>>>> If it still fails, please start cf-serverd with strace
>>>>>>>>>>>
>>>>>>>>>>> ```
>>>>>>>>>>> strace /var/cfengine/bin/cf-serverd -F
>>>>>>>>>>> ```
>>>>>>>>>>>
>>>>>>>>>>> Hopefully the policy fix resolves the issue for you.
>>>>>>>>>>>
>>>>>>>>>>> -Craig
>>>>>>>>>>> On Wednesday, July 7, 2021 at 12:36:05 PM UTC-5 Beto wrote:
>>>>>>>>>>>
>>>>>>>>>>>> I installed community 3.18.0 and masterfiles 3.18.0.
>>>>>>>>>>>> cf-serverd fails with the message in the subject and exits with rc 1,
>>>>>>>>>>>> which I can't find documented. No other errors appear in verbose output.
>>>>>>>>>>>>
>>>>>>>>>>>> Anyone got a clue? TIA
>>>>>>>>>>>>
>>>>>>>>>>>
--
You received this message because you are subscribed to the Google Groups "help-cfengine" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [email protected].
To view this discussion on the web visit https://groups.google.com/d/msgid/help-cfengine/9ea9b6ce-9313-4916-832a-5899a7adec51n%40googlegroups.com.