So, I just had to do this "abort, report, update" loop for around 8 times now to fix it.
It didn't resend the same tasks, but other ones.
Here's a (possibly uneducated) guess for you. I took a quick look at the tasks list but certainly not an 'in depth' one.
The host you mentioned in your initial message now shows nearly 400 aborted GW tasks. the earliest ones show as aborted on Dec 31st. The latest ones - Jan 3rd. As I scanned back from the latest, the dates were getting progressively earlier so that fits in quite nicely with your "around 8 times".
So here's the thing. If there are lots of tasks that the scheduler considers as 'lost', it sends them in batches of 12 at a time, irrespective of how many there are in total. If you abort and report just the batch of 12, it will happily accept them and then immediately send you another batch - for as long as necessary to deplete the entire stock of lost tasks that it has accumulated.
The best way to avoid this is to 'encourage' the scheduler to keep sending the lost tasks until its supply is exhausted. Then you can select the entire bunch and abort/report in a single transaction.
Hype wrote:
Don't know how they got lost because I always properly abort and report, but whatever.
It looks like the ones dated Dec 31st were not the complete list and somehow there were others remaining at the time the project was reset. However I'm glad that you now seem to have exhausted the supply of all those others :-).
So, I just had to do this
)
So, I just had to do this "abort, report, update" loop for around 8 times now to fix it.
It didn't resend the same tasks, but other ones. Don't know how they got lost because I always properly abort and report, but whatever.
Thanks for the help guys
Hype wrote: So, I just had
)
Here's a (possibly uneducated) guess for you. I took a quick look at the tasks list but certainly not an 'in depth' one.
The host you mentioned in your initial message now shows nearly 400 aborted GW tasks. the earliest ones show as aborted on Dec 31st. The latest ones - Jan 3rd. As I scanned back from the latest, the dates were getting progressively earlier so that fits in quite nicely with your "around 8 times".
So here's the thing. If there are lots of tasks that the scheduler considers as 'lost', it sends them in batches of 12 at a time, irrespective of how many there are in total. If you abort and report just the batch of 12, it will happily accept them and then immediately send you another batch - for as long as necessary to deplete the entire stock of lost tasks that it has accumulated.
The best way to avoid this is to 'encourage' the scheduler to keep sending the lost tasks until its supply is exhausted. Then you can select the entire bunch and abort/report in a single transaction.
It looks like the ones dated Dec 31st were not the complete list and somehow there were others remaining at the time the project was reset. However I'm glad that you now seem to have exhausted the supply of all those others :-).
Cheers,
Gary.