| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Smash Childhood Cancer
Thread: Why Is the Server Using Replication 3 on Some WUs? |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 13
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Project Name: Smash Childhood Cancer
----------------------------------------Created: 09/10/2019 17:25:27 Name: SCC1_0003451_OPN-b_D_13491 Minimum Quorum: 2 Replication: 3 If _0 and _1 are determined to be invalid, wouldn't just _2 and _3 be sufficient to make a valid work unit (which is quorum 2, replication 2) assuming both return as valid. Why is the _4 WU being sent? Are we doing more work than is necessary? [Edit 1 times, last edit by Doneske at Sep 11, 2019 9:16:48 PM] |
||
|
|
adriverhoef
Master Cruncher The Netherlands Joined: Apr 3, 2009 Post Count: 2155 Status: Offline Project Badges: 
|
There is suddenly a high rate of Invalids on several of my devices in the past hour!
----------------------------------------It's a mix of Valid and Invalid. Also, I see several WUs "Waiting to be sent". Result Name OS AVN Status Sent Time Due / Return Time CPUh Claimed/Gr.Result Name OS AVN Status Sent Time Due / Return Time CPUh Claimed/Gr.Result Name OS AVN Status Sent Time Due / Return Time CPUh Claimed/Gr.Result Name OS AVN Status Sent Time Due / Return Time CPUh Claimed/Gr.Result Name OS AVN Status Sent Time Due / Return Time CPUh Claimed/Gr.Result Name OS AVN Status Sent Time Due / Return Time CPUh Claimed/Gr.[Edit 2 times, last edit by adriverhoef at Sep 11, 2019 9:01:14 PM] |
||
|
|
UBT - JohnR
Cruncher Joined: Apr 30, 2006 Post Count: 35 Status: Offline Project Badges: 
|
The validator is not working correctly.
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Since having 2 invalids the supply has been drying up and the message for SCC which would normally show because I set the profile to limit the supply
----------------------------------------222 World Community Grid 9/11/2019 8:52:22 PM No tasks are available for Smash Childhood Cancer has disappeared... there's simply nothing coming for SCC. Of course because of the 2 invalid, any new work requires a wingman and all the completed work is being changed to pending verification, having lost reliability i.e. piles of sudden repair jobs being sent out, but where are the reliable hosts since all are losing reliability. edit: got some but the "No tasks are available" remains gone. [Edit 3 times, last edit by Former Member at Sep 11, 2019 9:48:15 PM] |
||
|
|
Stiwi
Advanced Cruncher Joined: May 19, 2012 Post Count: 75 Status: Offline Project Badges: 
|
I got also many invalids/ pending verification :(
|
||
|
|
JMrkvicka
Cruncher United States Joined: Jul 14, 2005 Post Count: 33 Status: Offline Project Badges: 
|
Something interesting is going on. I am set to crunch only SCC (HSTB is turned on but tasks for this project are as rare as honest politicians). I have averaged ~100-120 completed tasks per day. My average for Monday and Tuesday this week is ~2500 and it appears I am on track for another big day today. Are the huge number is 15 minutes tasks expected?
---------------------------------------- |
||
|
|
JMrkvicka
Cruncher United States Joined: Jul 14, 2005 Post Count: 33 Status: Offline Project Badges:
|
And now it appears I shot the golden goose!
----------------------------------------09/11/19 6:38:32 PM | World Community Grid | Reporting 1 completed tasks 09/11/19 6:38:32 PM | World Community Grid | Requesting new tasks for CPU 09/11/19 6:38:33 PM | World Community Grid | Scheduler request completed: got 0 new tasks 09/11/19 6:38:33 PM | World Community Grid | No tasks sent 09/11/19 6:38:33 PM | World Community Grid | No tasks are available for Help Stop TB 09/11/19 6:38:33 PM | World Community Grid | No tasks are available for the applications you have selected. 09/11/19 6:38:33 PM | World Community Grid | Tasks are committed to other platforms |
||
|
|
DrMason
Senior Cruncher Joined: Mar 16, 2007 Post Count: 153 Status: Offline Project Badges:
|
The OPN-b_C_, _D_, and _E_ sets are all very short tasks. That is probably why JMrkvicka is seeing surprisingly high results all of the sudden. That's not a problem necessarily.
----------------------------------------I can also report that I'm having a small, but significant, number of my returns labeled invalid. Around 11:22:15, from 13:08:50 to 13:13:43, from 16:07:25 to 16:32:35, and 18:00:36 to the present (18:54:32). There are many results that are valid, but I have gotten about 210 invalid results, so I'm slightly concerned as well, since I did not see this issue before. All my machines are definitely validated as returning valid results, and I haven't changed anything today. So, UBT - Johnr might be onto something with the validator.  |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7662 Status: Offline Project Badges:
|
Between 2 days ago and yesterday I had about 18000 units done and I see about 45 invalid units for those 2 days and today. I am on track to do about 8000 units today. That comes to about .173%. Granted that is higher than my normal 0% invalid, but I don't think it is necessarily bad considering the number of units processed. Given the number of work units in play there are bound to be a few errors. The invalids are coming from the machines which have processed the most units and with one exception are all from machines running Linux. Several machines show no invalids.
----------------------------------------I seem to recall on the single quorum projects, once the number of units which are either errors or invalids reach 5, the unit is pulled from further processing and the techs ( or scientists) decide if the unit is just bad or can be fixed and resubmitted. Because of the redundancies built into the BOINC framework, at this time it does not appear to be of great concern. Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I have 10 machines running that are using various Intel processors and 2 machines running AMD processors. I currently have about 250 WUs flagged invalid and 4200 flagged as Pending Verification. Sunday, I uploaded almost 60,000 WUs that resulted in 0 invalids and less than 15 pending verification. All of the WUs pending verification and/or marked invalid are from the Intel processors. That is not saying Intel is the problem as one of the AMD processors isn't running SCC1 as the project cannot keep the machine busy so it is exclusively on MIP1 and the other machine is an 8 core AMD processor that doesn't run that much. The interesting thing to me is I don't have a single anomaly prior to 18:52 UTC today on any machine. I don't think it is WU related as each machine was working on different WUs at the time the invalids started occurring. Did something happen on the server side at about 18:52 to cause things to start becoming invalid? I don't think it is a member problem as several members are reporting the same thing.
----------------------------------------[Edit 1 times, last edit by Doneske at Sep 12, 2019 1:14:43 AM] |
||
|
|
|