| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Active Research
Forum: OpenPandemics - COVID-19 Project
Thread: OpenPandemics - GPU Stress Test |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 781
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I wonder if the load balancers are being overloaded since the connections issues seem to be to any server (download/upload, forums, etc).
|
||
|
|
uplinger
Former World Community Grid Tech Joined: May 23, 2005 Post Count: 3952 Status: Offline Project Badges: 
|
We are working on the load balancer right now. :) It is quite busy at the moment.
Thanks, -Uplinger |
||
|
|
robertmiles
Senior Cruncher US Joined: Apr 16, 2008 Post Count: 445 Status: Offline Project Badges: 
|
Currently having BOINC up / download issues for WCG... I'm almost overloading my upload link, but no definite sign of also overloading the download side of the same link. Many ISPs make the download side of the link faster than the upload side, unless you pay for a commercial link that makes the upload side fast also. |
||
|
|
Ian-n-Steve C.
Senior Cruncher United States Joined: May 15, 2020 Post Count: 180 Status: Offline Project Badges: 
|
GPU app so host = GPU that's not how BOINC operates. host = the entire system CPU = processor GPU = coprocessor you can call it whatever you like yourself, but that's not the accepted terminology for the BOINC platform. . Whatever man. I’ll be sure and write that down. Your narcissism is starting to show. lol ok. I just correct misinformation when I see it. it's ok to be wrong.   EPYC 7V12 / [5] RTX A4000 EPYC 7B12 / [5] RTX 3080Ti + [2] RTX 2080Ti EPYC 7B12 / [6] RTX 3070Ti + [2] RTX 3060 [2] EPYC 7642 / [2] RTX 2080Ti |
||
|
|
Chooka
Cruncher Australia Joined: Jan 25, 2017 Post Count: 49 Status: Offline Project Badges:
|
Gee it's going to be tough to work out how many tasks to run consecutively when the run times are all over the shop.
----------------------------------------Looking at 1 of my pc's running 1 wu at a time, the wu's range from 3:19 - 6:03. I am running cpu work at the same time though so this might impact the consistency of the run times.   |
||
|
|
cjslman
Master Cruncher Mexico Joined: Nov 23, 2004 Post Count: 2082 Status: Offline Project Badges: 
|
Techs, might want to put a notification in the News that the stress test is going on because the site is slower than molasses going uphill during winter
---------------------------------------- (besides the other issues with download/upload, validation, etc). Luck with the testing  CJSL Crunching like Cap'n Crunch |
||
|
|
Ian-n-Steve C.
Senior Cruncher United States Joined: May 15, 2020 Post Count: 180 Status: Offline Project Badges:
|
my three multi-GPU systems, I'm only running 1x since I don't want to steal away so many threads from the CPU work. the two single-GPU systems, I'm running at 2x.
----------------------------------------EPYC 7V12 / [5] RTX A4000 EPYC 7B12 / [5] RTX 3080Ti + [2] RTX 2080Ti EPYC 7B12 / [6] RTX 3070Ti + [2] RTX 3060 [2] EPYC 7642 / [2] RTX 2080Ti |
||
|
|
uplinger
Former World Community Grid Tech Joined: May 23, 2005 Post Count: 3952 Status: Offline Project Badges:
|
Gee it's going to be tough to work out how many tasks to run consecutively when the run times are all over the shop. Looking at 1 of my pc's running 1 wu at a time, the wu's range from 3:19 - 6:03. I am running cpu work at the same time though so this might impact the consistency of the run times. The work unit runtimes vary because there are ways for it to stop early if it finds a good answer. This helps improve the efficiency of the work unit. Also, when we size them to start, it is just a complete estimate of how difficult they will be. On average the estimate works, but on a single work unit they can vary some. Thanks, -Uplinger |
||
|
|
spRocket
Senior Cruncher Joined: Mar 25, 2020 Post Count: 280 Status: Offline Project Badges: 
|
I've suspended the three ARP units in flight on my system, since more than once today I've had to restart BOINC when it started getting stupid about stuck downloads. Since ARP rarely checkpoints its work, that's a huge hit.
BOINC seems to be very talented at backing off a short while, then inevitably failing again and again on that same file to where it starts backing off 30+ minutes, even as other files merrily pass by back and forth. There doesn't seem to be any way to unstick the stuck jobs other than by restarting it or hoping my queue doesn't run dry. We're certainly getting our stress test. |
||
|
|
Grumpy Swede
Master Cruncher Svíþjóð Joined: Apr 10, 2020 Post Count: 2500 Status: Recently Active Project Badges:
|
Our dear friend the Validator, does not seem able to keep up with all reported work. The backlog of Pending Validation is growing even for tasks where I have no wingman.
Oh dear, how will I be able to continue living?  |
||
|
|
|