Welcome to MilkyWay@home

Posts by Wisesooth

1) Message boards : News : Server Maintenance (June 1st, 2023) (Message 75430)
Posted 2 Jun 2023 by Profile Wisesooth
Post:
Now that power maintenance is complete, could you do us people with Intel K processors a favor? Please make Modfit a separate category so we can exclude multicore tasks. I have to abort-by-user all 8-core tasks because they hang my systems.
FYI, the K processor has four real and four virtual cores. That works if, and only if, both the virtual core and the real core that manages it try to access the same instruction (e.g.: double-precision floating point) at the same time. When this happens, both the virtual and the real core go into a wait state forever. I have to manually suspend the task and renew it to release the wait states. This happens more than once before each 8-core task completes. If I abort the 8-core tasks by-user without selecting "no more tasks" first, the tasks I just aborted come back to me as new tasks. Getting BOINC, Microsoft, and/or Intel to fix this is useless.
2) Message boards : News : Server Downtime 3/21 1PM EST (Message 72148)
Posted 22 Mar 2022 by Profile Wisesooth
Post:
Pass this on to Tom.
I discovered a way to clear the multiple-core wait lock issue. Suspend the task to park it in cache. That clears both the real and virtual core's locked wait state.. Another task grabs the cores. Next, resume the task in the cache. That action puts it in a pending state. It finishes the task when it becomes active and finishes normally. Roger O
3) Message boards : News : New Poll Regarding GPU Application of N-Body (Message 71008)
Posted 23 Jul 2021 by Profile Wisesooth
Post:
There is an "anomaly" in the latest version of BOINC client's use of tasks that use the GPU option. It has something to do with the Intel GPU. It can hang a task indefinitely. Trod carefully before checking this out. As Redbeard the pirate might say if he were still alive, "Matey, ye be warned!"
4) Message boards : News : Where MilkyWay@home Was The Last Few Days (Message 70793)
Posted 17 May 2021 by Profile Wisesooth
Post:
The internet is being used as a weapon by at least three types of sources:
    Government predators(especially their military),
    Ideological paramilitary (e.g.: Islamic jihad), and
    Organized crime (e.g.: identity theft).


Unfortunately, Internet protocol technology did not pay due attention to abuse in its design. Firmware technology used in routers and motherboards ignored this threat, even though it was detected and known for most of the current decade (e.g.: trapdoor chip). Recent performance advances in computer chip technology and artificial intelligence have made current state-of-the-art encryption more vulnerable to decryption. Also, there is urgent need for more intensive information-sharing and coordinated research between public and private sector organizations (with international scope) to abate the risk to tolerable levels. I have more to share about this subject that is beyond the limitations of a comment. Some of the information is best shared privately. You know where to find me, Tom.

5) Message boards : News : Database Maintenance 9-4-2014 (Message 67773)
Posted 5 Sep 2018 by Profile Wisesooth
Post:
IMHO, the database server seems to be the weakest link in your system. A DBMS is the most processor and storage intense application in a system like this. RPI really needs a server with enough cores and solid-state memory to handle the throughput required to manage a grid computing environment, especially if the DBMS is enforcing referential integrity. Intel might give RPI some hardware help if they ask.
Yes, I know you think "Yes, we already know that." However, do the people with the purse know that? If they are not listening, maybe forwarding this message might get their attention. Overloaded servers breakdown at the most inconvenient times. The cost and time of RPI's most talented people should be considered in the total cost of ownership.
6) Message boards : News : Scheduled Server Maintenance 3/27 (Message 67279)
Posted 27 Mar 2018 by Profile Wisesooth
Post:
Hello Jake,
I am an Intel partner with history of server builds for clients. I do not have enough information to give you precise advice, so the following may not be applicable to your situation.
Processor sockets can sometimes be fixed with an alcohol swab if accidentally touched by human hands or gunked up with processor grease. Those contacts and the immediate neighborhood must be "squeaky clean." Also, processor grease is not a commodity item. I use processor grease with silver powder inside (sparingly) for servers destined for extreme use. Of course, broken or bent pins cannot be fixed. If this is the case, the motherboard is toast. When a motherboard needs replacement, they don't make them any more. That means the processors, RAM, power supply and other stuff needs replacement because it is not compatible with newer boards currently in production. Now that 14 nanometer technology is the new normal, nothing older than 3 years has the correct operating voltage. If you find a "legacy" replacement board, remember this. You don't know where it has been.
7) Message boards : News : Scheduled Server Maintenance 2/21 (Message 67131)
Posted 23 Feb 2018 by Profile Wisesooth
Post:
The database server app is generating database errors. It is not receiving work or responding to user-requested updates. This has been going on for hours.
8) Message boards : News : Repealing Net Neutrality Could Affect MilkyWay@home (Message 66884)
Posted 24 Dec 2017 by Profile Wisesooth
Post:
What does Canada have to do with U.S. net neutrality? I am curious.

BTW, repeal of net neutrality just transfers jurisdiction back to the Federal Trade Commission (FTC, not FCC). This agency is protected by the U.S. Constitution's enumerated power of Congress (not the Administrative Branch) to regulate interstate commerce.

The Net neutrality law was an unconstitutional power grab by the Administrative branch from the Congressional Branch of the U.S. government. The President of the United States (POTUS) justified the grab by claiming that Internet Service Providers (ISPs) were public utilities.

Did you all know that the National Broadcasting Company (NBC) owns Comcast? Indirectly, the FCC has reach into Comcast, Time Warner and Verizon. If an ISP misbehaves, the FTC could use the Clayton Act (an antitrust law) to cite the ISP for restraint of trade. Defense lawyers love antitrust laws. It is a cash cow that ruins an ISP's bottom line for years. Plaintiffs get triple damages if they win the case. ISPs know that.
9) Message boards : News : New Team Creation Requirement (1 Credit Awarded) (Message 66856)
Posted 15 Dec 2017 by Profile Wisesooth
Post:
You might consider reporting the more aggregious spammers to the FBI via spam@uce.gov
Also, ISPs subscribe to special services they use to block incoming URLs on their black list. If you get the URLs on their black list, they do not even get through the ISPs to your site.
10) Message boards : News : Repealing Net Neutrality Could Affect MilkyWay@home (Message 66855)
Posted 15 Dec 2017 by Profile Wisesooth
Post:
We might find a better solution to the problem with the anti-trust laws, particularly the Robinson-Pattman Act of 1933. This act deals with discrimination against the customer. This law has triple damages built within the statute and class actions are allowed.
11) Message boards : News : Repealing Net Neutrality Could Affect MilkyWay@home (Message 66841)
Posted 12 Dec 2017 by Profile Wisesooth
Post:
This net neutrality issue was never about neutrality. Instead, it was a step toward nationalization of the US Internet leading to a national firewall like China (PRC). If anyone thinks that the FCC is neutral about anything and above corruption, they need to study their history. The FCC is an agency of the US Administrative Branch led by the President, not Congress. That gives the President the power to monitor, control and censor electronic communication without oversight or accountability. It also provides an irresistible urge by Congress to tax its use, just like they did to the phone system. Human nature predictably cannot handle that much power.
We need a better solution that works with imperfect people at the helm. We are not even close to a solution to the problem because we have not properly defined the problem. A bureaucrat's only competition is another bureaucrat.
The surest way to make the new system cost more than the old is to pioneer the wrong thing. The surest way to lose in poker or blow a budget to smithereens is to throw good money after bad trying to fix an unworkable solution. The surest way to make a system fail is to allow a politician to design it.
Fairness and neutrality are like beauty. Its metrics lie within the opinion of the beholder.
As Redbeard might say. "Matey, ye be warned."
12) Message boards : News : Clearing the Queue (Message 66723)
Posted 21 Oct 2017 by Profile Wisesooth
Post:
"To err is human; to forgive is not company policy." Is that how it goes?

Let's plant our personal feelings aside and look at the problem. Once we discover what is wrong and what is right, let us help these budding academics instead of rejecting them as a person.

1. These points and "attaguy" badges are worth nothing but bragging rights. That stuff is for gamers, not us.
2. "No more work" will not help them clear the queue. It just postpones a solution to the problem.
3. The nature of the work they do is heuristic, not deterministic, which does not guarantee convergence. Third-order differential equations are difficult to solve. Tensors are even more difficult to solve. Let's cut these people some slack, shall we?
4. We have the option to abort what comes in by user. Let's do that.
5. Sidd and Jake, have you considered a "cleaner" project? This could initiate an algorithm to identify and compile a profile of tasks to kill and then automatically initiate an "abort by user" message. Those of us who want to help you clear the queue could add the cleaner project to their list of projects. That would clear the queue. When tasks stop coming in, shutdown the cleaner project.
13) Message boards : News : Bad Runs Over the Weekend (Message 66654)
Posted 19 Sep 2017 by Profile Wisesooth
Post:
BOINC upgraded to 7.8.2 at about the time this started happening. Coincidence?
14) Message boards : News : Server Maintenance (Message 66511)
Posted 3 Jul 2017 by Profile Wisesooth
Post:
The prior pose recorded 44 cores. I meant 4 cores. My keyboard stutters at times.
15) Message boards : News : Server Maintenance (Message 66510)
Posted 3 Jul 2017 by Profile Wisesooth
Post:
I caught the same error messages I found before the bundling remedy. The database server was overwhelmed again, even with bundling. I removed nbody from my active project list in my preferences to avoid complications. One of the high-volume projects (IBM's World Community) had a two-day outage. Those clients running both IBM and milkyway projects got a huge bump in milkyway runtime for a week because IBM gradually "trickled up" the uploaded work for at least three more days. Also, something is unstable in the nbody code. I had to abort a job that ran for over a day while it tied up 44 cores with no discernible progress. Hope this helps.
16) Message boards : News : New Release 1.64 (Message 66391)
Posted 6 May 2017 by Profile Wisesooth
Post:
I am getting lots of compute errors. Validate tasks from other users are having the same problem. Last I looked at my task results, I had 28 errors. I just suspended Milky Way project on all machines. Jake, you have my contact info. Let me know when this is fixed.
17) Message boards : News : Scheduled Maintenance Concluded (Message 65947)
Posted 23 Nov 2016 by Profile Wisesooth
Post:
I am not getting new work units. Did someone turn off the spigot?
18) Message boards : News : Updated Server Daemons and Libraries (Message 65350)
Posted 29 Sep 2016 by Profile Wisesooth
Post:
I Increased preferences about HD use from 40 to 80 and updated BOINC manager before I returned here to post this message.
BOINC Mgr reports:
Used by BOINC 15.91 MB Used by milkyway 9.9 MB Used by other programs 27.74 GB Free but not available to BOINC is 823.22 GB
This machine has 16 GB RAM and has a 1 TB Seagate hard drive. Uses Win 10 64-bit OS with Intel i7 processor. The only BOINC task running on this machine is milkyway at home. My profile currently shows over 50 tasks with computational errors. Other users are reporting the same errors on these tasks.

My other machines are running single CPU tasks with no errors on non-milkyway projects. Hope this helps you. BTW, have you considered the possibility that BOINC, not milkyway, is causing this problem? After all, the only project I know about on BOINC using all available threads on a single task is milkyway.
19) Message boards : News : Updated Server Daemons and Libraries (Message 65317)
Posted 27 Sep 2016 by Profile Wisesooth
Post:
My preferences exclude GPU work. My machines use Intel 6-gen processors with GPU on die. However, I am still getting computation errors on 8-thread tasks, but a lot less than before now. Other crunchers that run these tasks also abort with computation errors, so the issue is repeatable with up to three additional and different users. I am getting no computation errors from another BOINC project unrelated to milkyway.

This leads me to believe that the cause is data driven, not a code error, and in need of defensive programming to avoid the problem. If one or more subroutines (or threads) are sequentially bound, and the prerequisite thread does not complete before another thread needs to use its results, the consequences are predictable (all bad). The most common computational error is a zero-divide error. If the prerequisite thread is supposed to return an address location, it could produce a computation error if it references a location that is either protected or out of bounds.

I have not coded anything in years, but remember the pain. Hope this helps you to look in the right places.
20) Questions and Answers : Windows : Computation errors (Message 65067)
Posted 24 Aug 2016 by Profile Wisesooth
Post:
I just noticed that I have over 40 tasks aborted due to computation error. My other BOINC project shows no computation errors. These are not GPU errors. I am not overclocking my CPUs and have no thermal problems. Most but not all of the aborts are mt tasks. One of them took 54 hours before it aborted.


Next 20

©2024 Astroinformatics Group