killing all stuck rss aggregators:
for i in `sendcommand rwho |grep -i rss |sed "s;|.*;;" `; do sendcommand "TERM ${i}"; done
Does that have something to do with stuck aggregators? (Items falling out of the use table because the aggregators were stuck, perhaps?)
yes, saw these too.
My current plan is to disalow the tdap on the usetable in such a situation.
For shure there is also a need to find out whats wrong; to get some more of why & where aggregators were stuck the latest patches add more information to the rwho command.
Also, if that code has been touched at all, are we making sure that we update the usetable entry with the current timestamp if we see the item again?
yes, it should. however, if its blocked for some reason in first place, it never will ;-)
We need to catch that, and also the condition that i.e. the remote host goes away for a while.
whoops, broken it is right now. still searching.
ok, stable again.
new tiny webcit js hack: zapping duplicates so one can get rid of them in one click.
yea, you could upgrade again, however it seems as if still not all situations are fixed.
I'm working on it.