That's a sin almost, to abort a perfectly good result upload. Soft booting the problem device and modem/router or following this FAQ: When Client-Server Communications Started to be Troublesome http://www.worldcommunitygrid.org/forums/wcg/viewthread?thread=21569 fixes 99.9 out of 100.

It's interesting, because the file ending in _0_0 is one of the smaller parts, most usually the big _4 being the CEP2 trouble part.

PS: Think a file upload is aborted after 100 tries 14 days from first upload attempt. The BOINCTasks utility actually keeps count and shows that in the Transfer view so there is lots of time to get a fix. If not sorted, we can work to add some log flags to the client config to further diagnose... if having time and interest to dig in.

edit: correction to upload attempt expiration.

I've never heard of any limits on how many times a file-transfer can be tried, and can't find it in the code either...

But, there is a giveup for too much elapsed time since 1st. try. With older clients this is 14 days, while with v6.10.0 and later clients the time was extended to 90 days instead.

Will other, near-future projects run by the WCG be arranged in this way too; tasks issued directly from WCG to crunchers, but returned to elsewhere (Harvard in this case)?