[Digikam-users] Digikam memory usage

[Alberto Ferrante]

Dear Gilles,
at the end I tried running Digikam through gdb (report below).
Unfortunately it does not show much useful information: the program is
not really crashing, as I said it is terminated after having filled all
the available memory.
I discovered one thing: besides my desktop PC, I have a laptop, also
with Fedora 16 64 bit installed. On that machine Digikam behavior seems
to be different. The memory usage seems reasonable there. It is pretty
strange as both machines have pretty standard installations of Fedora,
both of them with the proprietary Nvidia drivers. Is there any library
in particular that I should check?
I am wondering if it could be anything related to the number of cores in
the CPU... How many threads digikam starts while doing an import?

Regards,
    Alberto

======================================================================
gdb output (the backtrace was not available as the program did not crash):
======================================================================
Starting program: /usr/bin/digikam
warning: "/usr/lib/debug/usr/lib64/libQt3Support.so.4.8.0.debug":
separate debug info file has no debug info
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib64/libthread_db.so.1".
[New Thread 0x7fffedcbf700 (LWP 22633)]
[New Thread 0x7fffed4be700 (LWP 22634)]
[New Thread 0x7fffe7fff700 (LWP 22635)]
[New Thread 0x7fffe77fe700 (LWP 22636)]
[Thread 0x7fffe77fe700 (LWP 22636) exited]
QSqlDatabasePrivate::removeDatabase: connection 'ConnectionTest' is
still in use, all queries will cease to work.
[Thread 0x7fffe7fff700 (LWP 22635) exited]
[New Thread 0x7fffe7fff700 (LWP 22638)]
[Thread 0x7fffe7fff700 (LWP 22638) exited]
[New Thread 0x7fffe7fff700 (LWP 22643)]
[New Thread 0x7fffe77fe700 (LWP 22644)]
[Thread 0x7fffe77fe700 (LWP 22644) exited]
[New Thread 0x7fffe77fe700 (LWP 22645)]
[New Thread 0x7fffcb7c7700 (LWP 22646)]
[Thread 0x7fffe77fe700 (LWP 22645) exited]
[Thread 0x7fffcb7c7700 (LWP 22646) exited]
[New Thread 0x7fffcb7c7700 (LWP 22654)]
[New Thread 0x7fffe77fe700 (LWP 22663)]
[New Thread 0x7fffcde98700 (LWP 22664)]
[New Thread 0x7fffcd697700 (LWP 22667)]
[New Thread 0x7fffc9c45700 (LWP 22675)]
Detaching after fork from child process 22676.
Detaching after fork from child process 22677.
[Thread 0x7fffcb7c7700 (LWP 22654) exited]
[Thread 0x7fffcd697700 (LWP 22667) exited]
[Thread 0x7fffe77fe700 (LWP 22663) exited]
[Thread 0x7fffcde98700 (LWP 22664) exited]
[New Thread 0x7fffcde98700 (LWP 22684)]
[Thread 0x7fffcde98700 (LWP 22684) exited]
[Thread 0x7fffc9c45700 (LWP 22675) exited]
[Thread 0x7fffed4be700 (LWP 22634) exited]
[Thread 0x7fffedcbf700 (LWP 22633) exited]
[Thread 0x7fffe7fff700 (LWP 22643) exited]

Program terminated with signal SIGKILL, Killed.
The program no longer exists.
======================================================================
======================================================================


On 11/15/2011 06:36 PM, Alberto Ferrante wrote:
> Gilles,
> I do not think I will be able to do that soon, it takes time and I do
> not have much at the moment. We anyway know already why it crashes: it
> fills up all the available memory (8GB of RAM +14 of swap) and it gets
> terminated. I monitored memory usage and this is what happens.
> I can tell that the memory use increases with the number of images. It
> is "acceptable" below 1-2'000 images; it becomes unacceptable above that
> threshold.
> 
> If you can, please try importing a folder with, say, 10'000 images...
> Are you able? How much memory is it using after the folder scan phase?

-- 
Home page: http://www.alari.ch/people/alberto
Photo galleries : http://albertoferrante.name
Public key: http://www.alari.ch/people/alberto/keys/yahoo.asc