[Nepomuk] [Kde-pim] Nepomuk feeder queuing problem

[Christian Mollekopf]

On Wed, Jan 18, 2012, at 07:48 PM, Milian Wolff wrote:
> On Wednesday 18 January 2012 01:16:01 Christian Mollekopf wrote:
> > On Tuesday 17 January 2012 17.37:19 Michael Jansen wrote:
> > > On Monday, January 16, 2012 06:41:33 PM Christian Mollekopf wrote:
> > > > Hey Sebastian,
> > > > 
> > > > I don't see the problem with this as this timer just kickstarts the
> > > > processing and can be called as many times as you want. The ItemQueue
> > > > does the actual feeding of the data. The processing of the ItemQueue is
> > > > triggered by the processItem() function which simply does nothing if
> > > > called repeatedly (mRunningJobs > 0). processItem() is also the function
> > > > which gets called if you call setOnline() or addItem(), but that
> > > > shouldn't hurt as far as I can see.
> > > > 
> > > > So there should be at maximum two nepomuk jobs running at a time (one
> > > > per ItemQueue).
> > > > 
> > > > Are you sure virtuoso is really going crazy? While indexing virtuoso
> > > > takes all the cpu power it can get, but that seems normal to me. Also
> > > > the indexing can take veeery long, so it's quite possible that users
> > > > think virtuoso just went crazy.
> > > 
> > > No. I am sometimes sitting besides my desktop working on my laptop and
> > > enjoy virtuoso going crazy and stopping by just listening into the fan.
> > > So i put a top up and glimpsed over to see what happens.
> > > 
> > > It will go to 80-180% cpu for some time, then go back to normal, nearly
> > > all
> > > of the time. It never stops doing that here. Usually i was able to find
> > > out
> > > 
> > >  the culprit by just looking at top. Another process was usually right
> > > 
> > > behind virtuoso in cpu usage.
> > > 
> > > We fixed most of these problems.
> > > 
> > > This time it is different. Only virtuoso does it. Some akonadi process
> > > seem
> > > to hover at 3-5% for me but not really sure if there is any relation.
> > > 
> > > I have no clue if it is related. But on session login i enjoy virtuoso,
> > > nepomukstorage, kontact and akonadi_nepomukfeeder go onto a minutes long
> > > cpu and io burning session. I plan to have a look during the next days
> > > what is the reason for that.
> > 
> > That virtuoso peaks at session start is the same here. However after a
> > minute or so everything goes back to normal.
> > 
> > Note that the feeder doesn't even really use a lot of cpu here when its full
> > speed indexing, the load is really generated on virtuoso. You should
> > however see if the feeder is active if you open up akonadiconsole (it would
> > print which collection it is currently indexing as status message).
> 
> Since I always have this issue (virtuoso hogging up one core constantly)
> I'd 
> like to offer my help. If you have any ideas on where I should look or
> what 
> kind of information I should provide, please tell me.
> 

Thanks, but unfortunately I'm lost myself. Today I also managed to
trigger virtuoso going crazy while I was testing kmail searching.
I'll tell you if I know how to debug this.

Cheers,
Christian

> Cheers
> -- 
> Milian Wolff
> mail at milianw.de
> http://milianw.de
> Email had 1 attachment:
> + signature.asc
>   1k (application/pgp-signature)