Why are we getting new N-Body work units when we have a lot of Validation Inconclusive work units that still need to be validated

Your question makes no sense. Presumably the ones you need validated require someone else to do the check. Why should that stop new ones being released?

We should focus on validating the work units that need to be validated first and the only work units that should be sent out now are the resend work units to validate each other's work.Peter Husker I really don't care on what you think.Oh by the way I hope World Community Grid comes back soon with enough CPU and GPU work units so I can leave this project

---

We should focus on validating the work units that need to be validated first and the only work units that should be sent out now are the resend work units to validate each other's work.Peter Husker I really don't care on what you think.

It's not what I think, it's simple logic. The tasks you have that aren't validated will have a wingman who already has that task downloaded. What do you want the server to do? Send out nothing leaving most of us idle until you get your tasks validated? I don't think you understand how Boinc works.

Oh by the way I hope World Community Grid comes back soon with enough CPU and GPU work units

They're taking months to move a server. They're incompetant. Expected live date is now 9th May as something broke again. They can't even get the website to work, let alone the work server. I guess Bob dropped it when he put it in his van.

so I can leave this project

You've been with it for 8 years....

oh, I figured it out. In order to change the size of the buffer pool I need to recompile 2 of the server binaries. That's not impossible, but it's a pain and may result in more problems. I might have some time this weekend I could try to do that.

oh, I figured it out. In order to change the size of the buffer pool I need to recompile 2 of the server binaries. That's not impossible, but it's a pain and may result in more problems. I might have some time this weekend I could try to do that.

Thanks again for your efforts. I can't believe Boinc requires that just to change a pool size! but I shouldn't be surprised really, they never cease to amaze me with weird stuff.

And the saga continues ..

And the saga continues ..

Where would be the fun if Boinc just worked? Without our constant interaction to fix things, it wouldn't be a hobby. Well, perhaps still a lot of messing around with a soldering iron to make the GPUs work because someone didn't invent the power connector properly so it can take 30 amps.

We should focus on validating the work units that need to be validated first and the only work units that should be sent out now are the resend work units to validate each other's work.Peter Husker I really don't care on what you think.Oh by the way I hope World Community Grid comes back soon with enough CPU and GPU work units so I can leave this project

-1
@Robert:
Just in case you don't know, but you really don't have to wait till WCG comes back online in order to leave MW !

We should focus on validating the work units that need to be validated first and the only work units that should be sent out now are the resend work units to validate each other's work.Peter Husker I really don't care on what you think.

It's not what I think, it's simple logic. The tasks you have that aren't validated will have a wingman who already has that task downloaded. What do you want the server to do? Send out nothing leaving most of us idle until you get your tasks validated? I don't think you understand how Boinc works.

Oh by the way I hope World Community Grid comes back soon with enough CPU and GPU work units

They're taking months to move a server. They're incompetant. Expected live date is now 9th May as something broke again. They can't even get the website to work, let alone the work server. I guess Bob dropped it when he put it in his van.

so I can leave this project

You've been with it for 8 years....

+1

Thanks Tom. Maybe you could ask on a Boinc forum?

There is a Mailing List just for Boinc Admins and Tom can ask on there at any time

We should focus on validating the work units that need to be validated first and the only work units that should be sent out now are the resend work units to validate each other's work.Peter Husker I really don't care on what you think.Oh by the way I hope World Community Grid comes back soon with enough CPU and GPU work units so I can leave this project

I'm sure they will appreciate your help too!!

[quote}

Oh by the way I hope World Community Grid comes back soon with enough CPU and GPU work units

They're taking months to move a server. They're incompetant. Expected live date is now 9th May as something broke again. They can't even get the website to work, let alone the work server. I guess Bob dropped it when he put it in his van. [/quote]

I can't believe they would even consider that, with modern day backups you can backup the Server today and reinstall it on the new Server tomorrow and be up and running within another day or so. For us users it's probably not even a website link change as all the changes would be internal to the Server.

I had high hopes that my N-Body validation inconclusive rate would have gone down a lot faster then its doing as i still have almost 14,700 to be validated now and i am hearing other people wish their N-Body work units were validated faster then what is happening now

---

I had high hopes that my N-Body validation inconclusive rate would have gone down a lot faster then its doing as i still have almost 14,700 to be validated now and i am hearing other people wish their N-Body work units were validated faster then what is happening now

Is that just previously done work or does that include work units from today or recent. Some of mine are now 15 days waiting to be validated/rejected, I haven’t done any for a week.

We should focus on validating the work units that need to be validated first and the only work units that should be sent out now are the resend work units to validate each other's work.Peter Husker I really don't care on what you think.Oh by the way I hope World Community Grid comes back soon with enough CPU and GPU work units so I can leave this project

I'm sure they will appreciate your help too!!

Except you're not allowed to complain over there. I've been permanently banned three times.

I can't believe they would even consider that, with modern day backups you can backup the Server today and reinstall it on the new Server tomorrow and be up and running within another day or so. For us users it's probably not even a website link change as all the changes would be internal to the Server.

AFAIK IBM provided the old servers and Krembil provide the new ones. So just a case of installing software and transferring data.

I had high hopes that my N-Body validation inconclusive rate would have gone down a lot faster then its doing as i still have almost 14,700 to be validated now and i am hearing other people wish their N-Body work units were validated faster then what is happening now

Is that just previously done work or does that include work units from today or recent. Some of mine are now 15 days waiting to be validated/rejected, I haven’t done any for a week.

Not sure what timescale this is over, but my tasks page shows 1 valid and 2370 inconclusive.

https://milkyway.cs.rpi.edu/milkyway/results.php?userid=167211&offset=0&show_names=0&state=0&appid=2

Seperation is validating normally.

A lot of them were done before we had the server problems and most of the work units were created on March 15 or even before that time

---

A lot of them were done before we had the server problems and most of the work units were created on March 15 or even before that time

Those you may have to just forget about until they do a full database purge, lots of tasks were 'lost' to the Server over this past year and other than go into each persons list and remove them or do the purge they are just 'there'.

A lot of them were done before we had the server problems and most of the work units were created on March 15 or even before that time

Those you may have to just forget about until they do a full database purge, lots of tasks were 'lost' to the Server over this past year and other than go into each persons list and remove them or do the purge they are just 'there'.

I take it that these are work units blocked on a result with "Don't need" status...

Some of those tasks have already unlocked and validated... The ones I have seen appeared to be for initial tasks that hadn't responded before Tom did the "purge"... The original client then happened to respond in some fashion (or the result timed out) after Tom did the "purge" and a new retry was issued in response to that event -- within a few days two successful results had been accumulated and validation happened. So they can be moved on! :-)

The transitioner maintains a "next transition time" which is usually based on time-outs of "in progress" results but which will get set to a "safety-net" value if there are no "in progress" results or if the safety-net value is less than the earliest result time-out value. (This does not apply if the task has a canonical result or has errors that render it unusable.) So in theory it will eventually hit the safety-net value and re-visit these tasks, just as it did for ones that "timed out" or replied after the purge. However, a lot of work units had first results returned before the purge -- those are currently stuck!

The minimum safety-net time in the code I looked at was 10 days, but there are situations where it could be longer; if it does try to get the work unit(s) moving again we may see some activity on or after 30th April. If not, Mikey may well be right that we just have to "live with it..."

Cheers - Al.

P.S. my nightmare about this is that if/when the transitioner does try to progress those work units there is a sudden huge surge in the number of available N-body tasks again!