Just done my daily check of the status page. Number of users in last 24 hours shows Linux only three behind the windows tasks despite the, "number of tasks in progress" for Windows being about seven times as many as for Linux. I am guessing it must mean a lot of those windows tasks are never going to be returned.

Dave -

I am sure you are correct. I checked my account and I have 5 that will never be returned. Some were because of computer crashes which required me to upgrade to LINUX from Windows XP..

I would think the database people would want to "purge" tasks that are more than one year old (or so). It would make database administration a lot easier. I know they would need permission from the scientists, but they could just tell the scientists to re-submit them if they really want them. The scientists are not going to get a result now anyway.

---

I would think the database people would want to "purge" tasks that are more than one year old (or so). It would make database administration a lot easier. I know they would need permission from the scientists, but they could just tell the scientists to re-submit them if they really want them. The scientists are not going to get a result now anyway.

Pretty sure, these days, results that old won't be any use because the scientists have done the work they need to using the results they have got back.

And someone has returned a WAH2RI task" How long is it since any of them were sent out???

Just done my daily check of the status page. Number of users in last 24 hours shows Linux only three behind the windows tasks despite the, "number of tasks in progress" for Windows being about seven times as many as for Linux. I am guessing it must mean a lot of those windows tasks are never going to be returned.

I’ve known that for years. Most of those “in progress” tasks have been lost in ways that the server never found out about and will never be completed. With no significant new work for Windows in months just about everyone with running Windows tasks has finished them.

---

And someone has returned a WAH2RI task" How long is it since any of them were sent out???

Looking at tasks for my longest continually running PC, finally retired last fall, the last batches issued for RI might have been Oct 2016. So 3.5 years ago.

And someone has returned a WAH2RI task" How long is it since any of them were sent out???

Looking at tasks for my longest continually running PC, finally retired last fall, the last batches issued for RI might have been Oct 2016. So 3.5 years ago.

Even my atom powered netbook would have completed those tasks in under six months!

And someone has returned a WAH2RI task" How long is it since any of them were sent out???

Looking at tasks for my longest continually running PC, finally retired last fall, the last batches issued for RI might have been Oct 2016. So 3.5 years ago.

Even my atom powered netbook would have completed those tasks in under six months!

Most likely someone reactivated a very old computer that had been sitting in storage for a few years. I supposed it would be possible to pull up the trickles and see if there is a 3+ year gap between 2 of them.

---

And someone has returned a WAH2RI task" How long is it since any of them were sent out???

Looking at tasks for my longest continually running PC, finally retired last fall, the last batches issued for RI might have been Oct 2016. So 3.5 years ago.

Even my atom powered netbook would have completed those tasks in under six months!

Most likely someone reactivated a very old computer that had been sitting in storage for a few years. I supposed it would be possible to pull up the trickles and see if there is a 3+ year gap between 2 of them.

You are probably right Jim. a computer old enough to take that long would be missing the SSE2 instruction set to run those tasks.

Looking at tasks for my longest continually running PC, finally retired last fall, the last batches issued for RI might have been Oct 2016. So 3.5 years ago.

And really don't know why we have these still on the status page? UK Met Office Coupled Model Full Resolution Ocean

A long time after the fact ...

I run Linux VMs on my Mac Mini. Earlier this week, one of the VMs, which had ten tasks on it, suffered a catastrophic failure (apparently caused by the Mac virus checker hanging the hardware). I was unable to restart the VM, so had to bin it.

There appears to be no way to notify CPDN that all incomplete/errored tasks for machine 1507532 need to be reallocated. It seems a waste to let them hang in the ether until the deadline, in September 2021.

https://www.cpdn.org/results.php?userid=2000635

(but not machine 1511336 - those are running fine).

Nick

That will happen automatically when they time out.
Or sooner when the researchers have enough data and close that batch.

And if some tasks in a batch don't get returned soon enough for the researchers, they can issue another batch with replacement tasks.

That will happen automatically when they time out.
Or sooner when the researchers have enough data and close that batch.

And if some tasks in a batch don't get returned soon enough for the researchers, they can issue another batch with replacement tasks.

Thanks.