That task appears to have some more detailed issues. It shows that it worked on structure _3 several times during the run. Even after reaching structure _4 which should, be definition, mean any future restart would be to a structure number higher than 3.

I have EMailed the Project Team asking about this. But in the meantime, it looks like that task probably started and ended many times. This shouldn't cause any difference in how it behaves, but I wanted to point out that it doesn't run as efficiently. I would suggest you review your preferences and set the checkbox to leave suspended tasks in memory ("virtual" memory actually, i.e. the swap space). It will allow a task to be suspended to run a task from another project for a while, and then when it comes back it can pick up all the details from where it left off, rather than have to restart back from the last checkpoint taken. Or perhaps you rebooted your machine several times while that last one was running?

---

Rosetta Moderator: Mod.Sense

I, of course, have the suspended tasks left in virtual memory. I believe I have indeed restarted the computer twice during run of those tasks, but I wouldn't think it would affect the effectiveness this greatly.

Guess 6 hours tasks were an overkill for me after all..

...but I wouldn't think it would affect the effectiveness this greatly.

No, I'm sorry I wasn't more clear. I think the reprocessing back to model _3 several times was the ultimate cause of the low credit. But when model 3 completed, it should have saved everything and any later restart should have been from model 4.

So, no, your machine reboots simply seem to have revealed another problem, which resulted in that task not getting much work done, which resulted in low credit. The tasks are capable of storing intermediate results, and restarting at that point. But for some types of tasks, the intermediate results are not stored very often. This is where keeping tasks in memory can help. Especially if running multiple projects on the machine. I suspect there must be a problem in how those partial results are stored or used to restart, at least with that specific type of task. Thanks for reporting it. That is the first step to getting it resolved.

---

Rosetta Moderator: Mod.Sense