[ltt-dev] [PATCH] Poll : introduce poll_wait_exclusive() new function

[Mathieu Desnoyers]

* Andrew McDermott (andrew.mcdermott at windriver.com) wrote:
> 
> Mathieu Desnoyers <compudj at krystal.dyndns.org> writes:
> 
> [...]
> 
> >> > Mathieu Desnoyers explained it cause following problem to LTTng.
> >> > 
> >> >    In LTTng, all lttd readers are polling all the available debugfs files
> >> >    for data. This is principally because the number of reader threads is
> >> >    user-defined and there are typical workloads where a single CPU is
> >> >    producing most of the tracing data and all other CPUs are idle,
> >> >    available to consume data. It therefore makes sense not to tie those
> >> >    threads to specific buffers. However, when the number of threads grows,
> >> >    we face a "thundering herd" problem where many threads can be woken up
> >> >    and put back to sleep, leaving only a single thread doing useful work.
> >> 
> >> Why do you need to have so many threads banging a single device/file?
> >> Have one (or any other very little number) puller thread(s), that 
> >> activates with chucks of pulled data the other processing threads. That 
> >> way there's no need for a new wakeup abstraction.
> >> 
> >> 
> >> 
> >> - Davide
> >
> > One of the key design rule of LTTng is to do not depend on such
> > system-wide data structures, or entity (e.g. single manager thread).
> > Everything is per-cpu, and it does scale very well.
> >
> > I wonder how badly the approach you propose can scale on large NUMA
> > systems, where having to synchronize everything through a single thread
> > might become an important point of contention, just due to the cacheline
> > bouncing and extra scheduler activity involved.
> 
> But at the end of the day these threads end up writing to the (possibly)
> single spindle.  Isn't that the biggest bottlneck here?
> 

Not if those threads are either

- analysing the data on-the-fly without exporting it to disk
- sending the data through more than one network card
- Writing data to multiple disks

There are therefore ways to improve scalability by adding more data
output paths. Therefore, I don't want to limit scalability due to the
inner design, so that if someone has the resources to send the
information out at great speed scaleably, he can.

Mathieu

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