[ltt-dev] To limit the size of trace data

[Mathieu Desnoyers]

* Takashi Nishiie (t-nishiie at np.css.fujitsu.com) wrote:
> Hi.
> 
> 　To limit the size of trace data, can you enhance it?　I want to
> cooperate in development if it is possible.
> 　LTTng is operated by the background until some problems occur,
> and to want to do is to discontinue tracing and to analyze trace
> data when the problem occurs. I want to evade the output of huge
> trace data by the time the phenomenon that wants to be analyzed
> is generated.
> 　For instance, it is a function of acquiring trace data of 2GB
> or less, and maintaining the backlog of the fourth generation from
> the new one.
> 　Or, trace data is usually maintained in the memory, and trace data
> is cut out from dump data acquired with kdump when kernel panic is
> generated.
> 

I think what you are trying to do is to use the "flight recorder mode" ?
When a specific condition is detected, you can stop tracing (from
userspace or from within the kernel) and then gather the information,
with lttd or from a kdump image.

I don't see what would have to be added to LTTng to achieve what you are
looking for ?

Mathieu

> Regards,
> Takashi Nishiie
> 
> 
> 
> 
> _______________________________________________
> ltt-dev mailing list
> ltt-dev at lists.casi.polymtl.ca
> http://lists.casi.polymtl.ca/cgi-bin/mailman/listinfo/ltt-dev
> 

-- 
Mathieu Desnoyers
OpenPGP key fingerprint: 8CD5 52C3 8E3C 4140 715F  BA06 3F25 A8FE 3BAE 9A68