I searched for this and couldn\'t find a solid answer. Some people speculated that 6 weeks of inactivity (no trickles) would result in the WU being reassigned.

Anyone have any evidence of this? I have 4 WU\'s that are 40% complete in queue waiting for another pair in front of them to finish. [long story]

I came across this \'6 week\' item back in 2004, but it may have been on the php forum, and from a long way back even then.

There is only one way to test the time limit. :)
But there is a simple solution, which I\'ve used myself several times: Stop the other models, then run the one(s) nearing the limit for one trickle. This will keep it (them) \"reserved\" for another n weeks.

And just because a model gets re-issued does not mean that you can\'t continue with it; it just means that the researchers will get a second opinion on that dataset.

---

Backups: Here

Thanks for the response Les,

And just because a model gets re-issued does not mean that you can\'t continue with it; it just means that the researchers will get a second opinion on that dataset.

I see. Just seems like a waste of months of CPU time, so I\'d prefer to avoid the above happening. I\'ll try your workaround though, hopefully that will do the trick.

Cheers,