http://aqua.dwavesys.com/

@$
*_*

---

Aqua is back online - recovering from last weeks outage.

There is (per the Aqua admins) a bug in the current BOINC server code that specifically affects Aqua work unit credit numbers. Some VERY large credit numbers are being generated.

This is likely to result in some stats site disruption. The Aqua admins are hopeful that the BOINC server code problem which specifically addresses these super large credit awards is resolved soon by a BOINC server code update, but until that happens (days/weeks) Aqua is going to dominate project stats sites and Aqua users are going to dominate individual stats reports as well.

From the Aqua project admins:

In preparation for a credit rollback, I've emptied the queues of all work. Please feel free to cancel any tasks that you may be running.

Please run tasks from other BOINC projects till we sort out this credit mess.

Looks like the primary problem with the server software is the broken 'credit new' scheme being pushed in the new code. It seems the idea of centralized dictation of credit awards is not only problematic from a 'freedom of choice' aspect, but also incorporates disfunctional code.

Aqua remains in stasis - perhaps still trying to figure out how to handle a rollback and restart. That's entirely speculative on my part -- no communications from the project since Friday.

AQUA problem before had announced a few minor repairs, Then came the problem of deleted files, I believe that about 15 days ago, we need to restart.

---

The Aqia credit problem first surfaced in June and was resolved by some code work around that allowed their existing credit scheme to continue functioning.

The problems they had at the end of the month resulted in them moving to a more recent server code iteration which (in line with Dave Anderson's central control of credits diktat) makes it increasingly difficult to go with project specific credit systems, resulted in the seriously flawed (and BOINC central code base) credits.

The project, seems to believe that the only acceptable solution is submitting to BOINC central -- though they are clearly conflicted about that since they realize that the BOINC central credit scheme is a MAJOR problem source.

Given that, and given that BOINC central is getting increasingly militant about controlling credit for all projects by their own server software, it appears that projects, whose applications are not readily comprehended by the BOINC server code developers, will increasingly either stop participating altogether or will figure out ways to 'fix' the server code so as not to impose the BOINC central credit schema on them.

I suspect we will see projects take that latter course.

I got this reply over in the Aqua message boards from one of the admins over there -- I think it provides a fair explanation of the how the project sees the credit issue.

There will be some folks who disagree with the project's position, just as there will be some folks who disagree with my vantage point, but I figured to post this here in the Projects area for Aqua to let any who track the posts over here know that the discussion CAN be civil.


BarryAZ, I agree with most of your points.

First, I am not defending CreditNew, or the whole idea of centralized credit assignment. To be honest, we consider the topic of credit to be secondary to the scientific goals of a project. But we know very well that we need our volunteers to be happy, and our volunteers are interested in receiving credit, so credit is important in that regard.

For AQUA, David Anderson has been a friend, not an enemy. We are software developers too and we know very well that working on a complex system like BOINC is hard. We would be surprised if there were never any bugs in such a system.

We may disagree with some of the recent changes in BOINC and were disappointed that we had to stop the work, but we don't consider disabling custom credit assignment as a personal insult (= we don't get emotional about it).

Our approach is to see if CreditNew can be made to work. Honestly, if there can be a standard and fair credit system then I am sure most project admins would be happy, because that is one less thing to worry about. Getting the scientific work done is the main concern of most project admins.

I know that the above may make some people angry, but please look at the big picture. BOINC credit is virtual. It is really just a token of appreciation for the help a volunteer has provided to a project. The most important thing here is that the amount of granted credit fairly represent the amount of help.

In the previous iteration of the outage thread I posted this:

"I understand -- the admin over at Aqua is aware that this can be done, but doesn't want to have DA get upset with Aqua by 'hacking' the server releases to avoid submission to the NewCredit diktat"

Based on further back and forth with the admin over at Aqua, I realize that this is not the position of the project.

That is, yes, they are aware that they could do a 'hack' to allow them to avoid the CreditNew scheme. But the reason they don't want to do this is that they feel that the CreditNew concept is one they support. The project does realize this is one of DA's special concepts, but it is not 'concern for upsetting DA' that motivates them to use CreditNew.

There are other projects which for various reasons don't (or can't) use CreditNew ;in some cases it is 'don't' because they are perhaps averse to central credit control for various reasons; it some cases it is 'can't' because the various versions of CreditNew don't accurately and reliably reflect the performance of their applications (though none that I aware of is as seriously off the mark as CreditNew for Aqua was prior to Aqua halting for a fix on CreditNew).

Be that as it may, again, the reason that Aqua is willing to work with CreditNew is because they conceptually support the centralized credit concept and would rather work toward make it work for Aqua (even with a temporary shut down), than work around CreditNew using an alternative approach.

Aqua is now 9 days into dealing the the NewCredit problem -- no resolution as of yet.

Also, (supposition only) it appears that the folks communicating about Aqua are (like many in BOINC projects) juggling a fair number of life/work priorities.

I make this supposition based on the pattern of communications seen on the Aqua message boards from the project. There tends not to be any project communication there on the weekends, nor during the mornings on weekdays.

That shouldn't be construed as anything but folks there dealing with other priorities -- as we all do.

Aqua has rolled back the mega credits -- the clean up should rattle into the stats sites over the next few days bringing things back into balance

From the long time admin over at Aqua:

Hi all,

Just a note to let you know that I am no longer with D-Wave, and this will be my last post as AQUA's admin. And before you ask: I don't know what will become of AQUA.

Thank you all for your help during the previous years!

-Kamran

Home-page announcement at AQUA@home

Thank-you all
Dear volunteers,

Your contribution over the last 3 years to the AQUA@home project was critical for us to perform immense computations necessary for the advancement of our science and technology. We have published several scientific papers reporting the results of those computations, including our recent publication in Nature. We made sure to acknowledge your generous contributions in those papers, as your help was vital. At this time, however, we no longer have large computations requiring the computing power of AQUA@home. We therefore have come to the decision to shut down AQUA for the forseeable future. Thank-you again for the selfless devotion of your time, patience, and computing resources to our effort to progress the field of quantum computing.

Sincerely,
Neil Dickson

Aqua it is the end.

Consistent with the announcement of Aqua's 'suspsension' for the 'foreseeable future' as of Friday (Sept. 9) the home page has been shut down.

Sorry to hear this. I was just about to experiment with it as a known good multi-threaded BOINC application on various Operating Systems.

:-(

Anyway:- what other CPU multi-thread BOINC applications are there? I have tried to find but there seems to be no categorisation by this criteria?

Thanks, Ray

---

The difference between 0 and 1 is greater than the difference between 1 and 1,000,000,000

Ray, I don't know of any at this point -- perhaps the folks at Dwave would be interested in selling the core approach they have used (they are a for profit business after all).

Anyway:- what other CPU multi-thread BOINC applications are there? I have tried to find but there seems to be no categorisation by this criteria?

Thanks, Ray

Isn't there a multi-threaded Milkyway@home CPU application?

Gruß,
Gundolf

Try http://www.renderfarm.fi/

It is now multithreading operational for the rendering clients, supported at the BOINC side.

Cheers,
Philip,
from Hong Kong