On Thu, Jul 18, 2013 at 9:34 AM, Erik Steffl <[email protected]> wrote:

> On 07/17/2013 10:01 PM, Rainer Gerhards wrote:
>
>> On Thu, Jul 18, 2013 at 2:43 AM, Erik Steffl <[email protected]> wrote:
>>
>>  On 07/16/2013 09:28 PM, David Lang wrote:
>>>
>>>  take a look at 
>>> http://www.rsyslog.com/doc/****node32.html<http://www.rsyslog.com/doc/**node32.html>
>>>> <http://www.**rsyslog.com/doc/node32.html<http://www.rsyslog.com/doc/node32.html>
>>>> >**and see if it
>>>> answers your questions.
>>>>
>>>>
>>>    since the queue is specified inside action does that mean that each
>>> action needs its own queue?
>>>
>>
>>
>> Yes, and thanks to the new config format this is now clear ;)
>>
>>
>>  E.g. I have two actions that forward messages to remote log aggregator
>>> and
>>> would like them to be assisted by the same disk queue (so that I can
>>> specify one queue.maxdiskspace). Is that possible?
>>>
>>>
>>>  No ... more below
>>
>>     action(
>>>      type="omfwd"
>>>      template="some template"
>>>      queue.filename="/var/log/****myFile"
>>>      queue.maxdiskspace="****4294967296"
>>>
>>>      ...
>>>    )
>>>    ...
>>>    action(
>>>      type="omfwd"
>>>      template="a different template"
>>>      queue.filename="/var/log/****myFile"
>>>      queue.maxdiskspace="****4294967296"
>>>
>>>      ...
>>>    )
>>>
>>>    Would this create one disk based queue with filename /var/log/myFile
>>> that both actions would use or would this just mess things up?
>>>
>>>
>> Old version messed things ab, current version rename the second queue file
>> and emit a warning message. So you'll end up with two seperate queues in
>> any case.
>>
>> Side-note: I guess you used omfile just as example, but one never news who
>> will dig this up via google: it is absolutely counter-productive to use
>> disk queus with files, as writing the actual output files is much faster
>> (10 to 20 times) than writing a disk queue. So don't do that. Disk queues
>> are for potentially slow outputs (databases, remote systems...).
>>
>
>   I am a bit confused, I use omfwd in the examples above (left out the
> rest of omfwd args for brevity), just added the queue.filename and
> queue.maxdiskspace to make sure I understand it corerectly. Do my examples
> look correct?
>

My fault. I just saw /var/log and my mind switched to omfile, ignoring the
facts. Sorry for the confusion. Yes, this sounds right (except that most
folks usually do not put spool files in /var/log, what probably is a good
idea).

Rainer

>
>   and yes, we plan to use the disk based queues so that the messages are
> not lost if the remote log aggregator goes down or there is a network
> problem or something like that...
>
>
>     Alternatively is there a way to specify total limit for all disk based
>>> queues? (didn't find anything so I guess not but hope never dies :)
>>>
>>>
>>>  No, that would required an immense amount of synchronization with the
>> current implementation, which makes this infeasable.
>>
>> Background info: the current queue system is stricly sequential. To do
>> what
>> you ask for would require a totally different subsystem, something along
>> the lines of the OS page file system. Actually, I think such a thing to
>> have would be a big plus. However, it is a very major task to do, I'd say
>> 3+ month of work. I would need external funding /sponsorship for such a
>> big
>> task (at least a larger part of it).
>>
>> A work-around may be to put things into their own ruleset and create a
>> ruleset queue. However, this means if one of the actions block, the other
>> are not executed until that block is solved. May be acceptable based on
>> the
>> use case.
>>
>> Sorry I have no better answer.
>>
>
>   ok, I'll just make sure there is enough space for all the queues (only
> have two so not a huge deal).
>
>   thanks!
>
>         erik
>
>
>> Rainer
>>
>>    thanks!
>>
>>>
>>>          erik
>>>
>>>
>>>
>>>  David Lang
>>>>
>>>> On Tue, 16 Jul 2013, Erik Steffl wrote:
>>>>
>>>>    Action definition:
>>>>
>>>>>
>>>>>   action(
>>>>>     type="omfwd"
>>>>>     target="somehost.com"
>>>>>     port="5140"
>>>>>     protocol="tcp"
>>>>>     template="json"
>>>>>   )
>>>>>
>>>>>   How to make this action use disk assisted queue? I did read
>>>>> http://www.rsyslog.com/doc/****omfwd.html<http://www.rsyslog.com/doc/**omfwd.html>
>>>>> <http://www.rsyslog.**com/doc/omfwd.html<http://www.rsyslog.com/doc/omfwd.html>
>>>>> >and
>>>>> http://www.rsyslog.com/doc/****queues.html<http://www.rsyslog.com/doc/**queues.html>
>>>>> <http://www.**rsyslog.com/doc/queues.html<http://www.rsyslog.com/doc/queues.html>
>>>>> >
>>>>>
>>>>>
>>>>>   Section 'Disk-Assisted Memory Queues' says 'If a disk queue name is
>>>>> defined for in-memory queues (via $<object>QueueFileName), they
>>>>> automatically become "disk-assisted" (DA)'. However I also see at
>>>>> http://www.rsyslog.com/doc/****rsyslog_conf_actions.html<http://www.rsyslog.com/doc/**rsyslog_conf_actions.html>
>>>>> <http**://www.rsyslog.com/doc/**rsyslog_conf_actions.html<http://www.rsyslog.com/doc/rsyslog_conf_actions.html>>that
>>>>> legacy
>>>>>
>>>>> format does not affect RainerScript actions (which is what I use).
>>>>>
>>>>>   Or is it going to automatically store it in a directory specified by
>>>>> this: $WorkDirectory /var/spool/rsyslog and I don't have to worry
>>>>> about it?
>>>>>
>>>>>   thanks!
>>>>>
>>>>>      erik
>>>>> ______________________________****_________________
>>>>> rsyslog mailing list
>>>>> http://lists.adiscon.net/****mailman/listinfo/rsyslog<http://lists.adiscon.net/**mailman/listinfo/rsyslog>
>>>>> <http:**//lists.adiscon.net/mailman/**listinfo/rsyslog<http://lists.adiscon.net/mailman/listinfo/rsyslog>
>>>>> >
>>>>> http://www.rsyslog.com/****professional-services/<http://www.rsyslog.com/**professional-services/>
>>>>> <http://**www.rsyslog.com/professional-**services/<http://www.rsyslog.com/professional-services/>
>>>>> >
>>>>>
>>>>> What's up with rsyslog? Follow https://twitter.com/rgerhards
>>>>> NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a
>>>>> myriad of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST
>>>>> if you DON'T LIKE THAT.
>>>>>
>>>>>   ______________________________****_________________
>>>>>
>>>> rsyslog mailing list
>>>> http://lists.adiscon.net/****mailman/listinfo/rsyslog<http://lists.adiscon.net/**mailman/listinfo/rsyslog>
>>>> <http:**//lists.adiscon.net/mailman/**listinfo/rsyslog<http://lists.adiscon.net/mailman/listinfo/rsyslog>
>>>> >
>>>> http://www.rsyslog.com/****professional-services/<http://www.rsyslog.com/**professional-services/>
>>>> <http://**www.rsyslog.com/professional-**services/<http://www.rsyslog.com/professional-services/>
>>>> >
>>>>
>>>> What's up with rsyslog? Follow https://twitter.com/rgerhards
>>>> NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad
>>>> of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you
>>>> DON'T LIKE THAT.
>>>>
>>>>
>>> ______________________________****_________________
>>> rsyslog mailing list
>>> http://lists.adiscon.net/****mailman/listinfo/rsyslog<http://lists.adiscon.net/**mailman/listinfo/rsyslog>
>>> <http:**//lists.adiscon.net/mailman/**listinfo/rsyslog<http://lists.adiscon.net/mailman/listinfo/rsyslog>
>>> >
>>> http://www.rsyslog.com/****professional-services/<http://www.rsyslog.com/**professional-services/>
>>> <http://**www.rsyslog.com/professional-**services/<http://www.rsyslog.com/professional-services/>
>>> >
>>>
>>> What's up with rsyslog? Follow https://twitter.com/rgerhards
>>> NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad
>>> of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you
>>> DON'T LIKE THAT.
>>>
>>>  ______________________________**_________________
>> rsyslog mailing list
>> http://lists.adiscon.net/**mailman/listinfo/rsyslog<http://lists.adiscon.net/mailman/listinfo/rsyslog>
>> http://www.rsyslog.com/**professional-services/<http://www.rsyslog.com/professional-services/>
>> What's up with rsyslog? Follow https://twitter.com/rgerhards
>> NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad
>> of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you
>> DON'T LIKE THAT.
>>
>>
> ______________________________**_________________
> rsyslog mailing list
> http://lists.adiscon.net/**mailman/listinfo/rsyslog<http://lists.adiscon.net/mailman/listinfo/rsyslog>
> http://www.rsyslog.com/**professional-services/<http://www.rsyslog.com/professional-services/>
> What's up with rsyslog? Follow https://twitter.com/rgerhards
> NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad
> of sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you
> DON'T LIKE THAT.
>
_______________________________________________
rsyslog mailing list
http://lists.adiscon.net/mailman/listinfo/rsyslog
http://www.rsyslog.com/professional-services/
What's up with rsyslog? Follow https://twitter.com/rgerhards
NOTE WELL: This is a PUBLIC mailing list, posts are ARCHIVED by a myriad of 
sites beyond our control. PLEASE UNSUBSCRIBE and DO NOT POST if you DON'T LIKE 
THAT.

Reply via email to