So what you're saying is that it would be awesome if somebody at RU could
post details of that for the list.


On Mon, May 3, 2010 at 12:37 PM, Scott Battaglia
<[email protected]>wrote:

> Yeah, its pretty trivial.  Since I don't work at RU, I can't steal their's
> to show it to you :-)
>
>
> On Mon, May 3, 2010 at 3:31 PM, Patrick Berry <[email protected]> wrote:
>
>> Understandable.  I was still running into issues with Tomcat not shutting
>> down cleanly when I used a datastore, which is why I went with the
>> file-based logging.  I suppose it would fairly trivial to create my own
>> class to do everything in one line, yes?
>>
>> Pat
>>
>> On Mon, May 3, 2010 at 11:39 AM, Scott Battaglia <
>> [email protected]> wrote:
>>
>>> We should be clear that that logger is not intended for production
>>> usage.  Its more for testing (though we never documented that explicitly).
>>> The reason its multiple lines is so that you can easily tell what's going on
>>> during your development phase.
>>>
>>> At Rutgers, we had written a one-liner that made it easy for us to use
>>> grep to find  anything we needed ;-)  Most people store it in a database.
>>>
>>>
>>> On Mon, May 3, 2010 at 2:25 PM, Marvin Addison <[email protected]
>>> > wrote:
>>>
>>>> > Would it be odd to see this kind of file logging in Inspektr?
>>>>
>>>> No.  This kind of thing can happen frequently when you have multiple
>>>> threads competing for IO on the same logger file.  In the case of CAS
>>>> you have Tomcat's worker thread pool handling multiple authentication
>>>> requests, each of which causes inspektr to write several sequential
>>>> log info messages constituting the blocks you quoted.  The only way to
>>>> prevent those audit blocks from being interleaved in that case is to
>>>> use a mutex in Slf4jLoggingAuditTrailManager#record(); that way other
>>>> threads would block until the complete set of info messages is written
>>>> by the current thread.
>>>>
>>>> I'd encourage you to open an Inspektr issue for this,
>>>> http://github.com/dima767/inspektr/issues/, and see what shakes out.
>>>> I think it's entirely reasonable to support a log parsing script use
>>>> case, which would clearly break at present as you noted.
>>>>
>>>> M
>>>>
>>>> --
>>>> You are currently subscribed to [email protected] as:
>>>> [email protected]
>>>>
>>>> To unsubscribe, change settings or access archives, see
>>>> http://www.ja-sig.org/wiki/display/JSG/cas-dev
>>>>
>>>
>>> --
>>> You are currently subscribed to [email protected] as: [email protected]
>>>
>>>
>>>
>>> To unsubscribe, change settings or access archives, see 
>>> http://www.ja-sig.org/wiki/display/JSG/cas-dev
>>>
>>>
>> --
>>
>> You are currently subscribed to [email protected] as: 
>> [email protected]
>>
>>
>>
>> To unsubscribe, change settings or access archives, see 
>> http://www.ja-sig.org/wiki/display/JSG/cas-dev
>>
>>
> --
> You are currently subscribed to [email protected] as: [email protected]
> To unsubscribe, change settings or access archives, see 
> http://www.ja-sig.org/wiki/display/JSG/cas-dev
>
>

-- 
You are currently subscribed to [email protected] as: 
[email protected]
To unsubscribe, change settings or access archives, see 
http://www.ja-sig.org/wiki/display/JSG/cas-dev

Reply via email to