Since the original bug is still present (but nobody cared to update the version number), I'm cloning that bug and cleaning up a bit.

+++ This bug was initially created as a clone of Bug #1047550 +++

Description of problem: When reporting a crash, ABRT informs me that the stack trace is too large for the retrace server, shows me the size of the stack trace and the max size allowed by the retrace server, and asks if I want to generate a stack trace locally. I say Yes, acknowledging that it might download a large amount of data. Then ABRT pops up two identical dialogs claiming that the stack trace does not exist.  At the bottom of my journal I see:

Dec 31 11:01:33 victory-road gnome-session[1661]: '/var/tmp/abrt/ccpp-2013-12-31-10:50:45-2136' does not exist
Dec 31 11:01:39 victory-road gnome-session[1661]: '/var/tmp/abrt/ccpp-2013-12-31-10:50:45-2136' does not exist

Scrolling up I see:

Dec 31 10:50:44 victory-road kernel: WebKitWebProces[2136]: segfault at a ip 00007f240c33d0a3 sp 00007fffae0a9ce0 error 4
Dec 31 10:51:05 victory-road abrt-hook-ccpp[3098]: Saved core dump of pid 2136 (/usr/libexec/WebKitWebProcess) to /var/tmp/abrt/ccpp-2013-12-31-10:50:
Dec 31 10:51:06 victory-road abrt-hook-ccpp[3098]: /var/tmp/abrt is 2424106393 bytes (more than 1279MiB), deleting 'ccpp-2013-12-30-20:25:47-27664'
Dec 31 10:51:06 victory-road abrt-hook-ccpp[3098]: /var/tmp/abrt is 2214642705 bytes (more than 1279MiB), deleting 'ccpp-2013-12-30-20:26:12-27966'
Dec 31 10:51:06 victory-road abrt-hook-ccpp[3098]: /var/tmp/abrt is 2024150614 bytes (more than 1279MiB), deleting 'ccpp-2013-12-23-13:13:28-2416'
Dec 31 10:51:06 victory-road abrt-hook-ccpp[3098]: /var/tmp/abrt is 1915054564 bytes (more than 1279MiB), deleting 'ccpp-2013-12-22-18:53:20-9041'
Dec 31 10:51:06 victory-road abrt-hook-ccpp[3098]: /var/tmp/abrt is 1754052640 bytes (more than 1279MiB), deleting 'ccpp-2013-12-30-10:53:59-2356'
Dec 31 10:51:06 victory-road abrt-hook-ccpp[3098]: /var/tmp/abrt is 1624407229 bytes (more than 1279MiB), deleting 'ccpp-2013-12-23-14:30:14-3846'
Dec 31 10:51:06 victory-road abrt-hook-ccpp[3098]: /var/tmp/abrt is 1508828082 bytes (more than 1279MiB), deleting 'ccpp-2013-12-25-22:00:59-2643'
Dec 31 10:51:06 victory-road abrt-hook-ccpp[3098]: /var/tmp/abrt is 1408447330 bytes (more than 1279MiB), deleting 'ccpp-2013-12-21-09:59:27-2623'
Dec 31 10:51:06 victory-road abrt-hook-ccpp[3098]: /var/tmp/abrt is 1348246420 bytes (more than 1279MiB), deleting 'ccpp-2013-12-22-21:15:38-18167'
Dec 31 10:51:09 victory-road abrt-server[3105]: Generating core_backtrace
Dec 31 10:51:09 victory-road abrt-server[3105]: Generating backtrace
Dec 31 10:51:15 victory-road gdb[3130]: detected unhandled Python exception
Dec 31 10:51:27 victory-road gdb[3179]: detected unhandled Python exception
Dec 31 10:51:43 victory-road dbus-daemon[900]: dbus[900]: [system] Activating service name='org.freedesktop.problems' (using servicehelper)
Dec 31 10:51:43 victory-road dbus-daemon[900]: dbus[900]: [system] Successfully activated service 'org.freedesktop.problems'
Dec 31 10:51:43 victory-road dbus[900]: [system] Activating service name='org.freedesktop.problems' (using servicehelper)
Dec 31 10:51:43 victory-road dbus[900]: [system] Successfully activated service 'org.freedesktop.problems'
Dec 31 11:00:50 victory-road kernel: traps: WebKitWebProces[3108] trap int3 ip:7f5c37ffe58a sp:7fffefec6070 error:0
Dec 31 11:00:55 victory-road abrt-hook-ccpp[3627]: Saved core dump of pid 3108 (/usr/libexec/WebKitWebProcess) to /var/tmp/abrt/ccpp-2013-12-31-11:00:
Dec 31 11:00:55 victory-road abrt-hook-ccpp[3627]: /var/tmp/abrt is 1627161331 bytes (more than 1279MiB), deleting 'ccpp-2013-12-31-10:50:45-2136'

Looks like ABRT is deleting old core dumps before saving a new one. It should at least not delete one that it's currently reporting a bug against.

Version-Release number of selected component (if applicable):

abrt-2.1.10-1.fc20
gnome-abrt-0.3.4-1.fc20
libreport-2.1.10-1.fc20
satyr-0.12-1.fc20

Additional info:

--- Additional comment from Michael Catanzaro on 2015-05-29 09:26:17 EDT ---

I think this has not been fixed, moving to F22.

--- Additional comment from Christian Stadelmann on 2015-09-24 10:24:27 EDT ---

Still present on F23.

--- Additional comment from Christian Stadelmann on 2017-11-23 07:41:49 EST ---

This issue is still present on Fedora 26.