| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Beta Testing
Forum: Beta Test Support Forum
Thread: new CEP2 beta started |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 32
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Wouldn't it be useful to adopt this n/2 cores feature for production? No manual check for too many CEP2 WUs running concurrently would be necessary anymore! Let me see.... Adopt a rule reducing invalid results by limiting the number of concurrent jobs OR allow crunchers to run as many WU as they want, creating multiple errors causing the system to generate repair WU that then error out causing further repair WU............. Thinking..... thinking...... |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Guys, I don't get my bandwidth above 114, I tried the HPF2 trick but somehow it doesn't work anymore. Is there another way to bypass that crap-limit?
----------------------------------------[Edit 1 times, last edit by Former Member at Sep 22, 2010 12:53:28 PM] |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
The limit is there for a reason... to protect those that are on real dial up / low bandwidth. Try 22MB upload on a 56kb modem. Get a series HCC jobs. They're rather short in run time but big enough to sway bwdown for this beta (if there's still work).
----------------------------------------By the time we get to production there will be a work around in place... only for 6.10 clients though so an upgrade is needed. Keep eye out on the member announcements so you get the right version.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The limit is probably not there to limit someone who could get 700kb downstream.
HCC don't work for me as I do normally only crunch HCC jobs and according to that Speed-O-Meter I do only get 100Kb downstream with it. |
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
Don't know, my duo client runs C4CW/HFCC/HCMD2 mix since a few days and got a Beta task and the control value indicates 341KB.
----------------------------------------<bwdown>341054.749134</bwdown> What ISP's tell and what is real at the wallsocket is not quite the same. Our contract says 20MB/1MB. On good days we get about 8-10MB from Rome (about 165 km) and from WCG (Toronto) mostly about 4MB on Speed-Test. Sometimes when downloading a series of larger jobs the client actually manages to hit a bwdown of 3MB. But, what we discuss here is mostly academic since the techs are working on the announced changes. cheers PS: Getting patches from the Linux shadow distributor in Italy opposed blows all stops... that does go awesome fast... works multithreaded.
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
littlepeaks
Veteran Cruncher USA Joined: Apr 28, 2007 Post Count: 748 Status: Offline Project Badges: 
|
I got one of these beta tests on my (one and only PC) Windows XP Pentium 4 PC yesterday. Ran exactly to the 12 hour cutoff and is in PV now. Amazingly, I got ANOTHER one, which is crunching now. Memory usage is high -- about 70,000 K average -- competes with Seamonkey, my browser, for highest memory usage.
|
||
|
|
Sekerob
Ace Cruncher Joined: Jul 24, 2005 Post Count: 20043 Status: Offline |
They are still to be had... increased the cache 0.1 day and it immediately fetched a new one n / 2 cores.
----------------------------------------22/09/2010 19:47:19 World Community Grid Finished download of E200499_A.25.C18H10N4S2Se.178.1.zip Starting manually... you never know if tomorrow there's more being injected into the feeder :>)
WCG
Please help to make the Forums an enjoyable experience for All! |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
To all guys who have problems with bandwidth:
This did the trick for me: http://www.worldcommunitygrid.org/forums/wcg/...ad,29293_offset,10#284849 |
||
|
|
rilian
Veteran Cruncher Ukraine - we rule! Joined: Jun 17, 2007 Post Count: 1452 Status: Offline Project Badges: 
|
Was lucky to get 3 another WUs this evening
----------------------------------------they start very strange. at first process wcgrid_beta11_6.25_windows_intelx86 hangs with it's child wcgrid_beta11_qchem_6.25_windows_intelx86 idle like 10 or more minutes, and then qchem starts crunching ... ps: crunching in Mac/Parallels virtual machine, and when this Wu starting, everything in VM hangs, like hangs with usual disk i/o hope this help somehow ---------------------------------------- [Edit 1 times, last edit by rilian at Sep 22, 2010 7:21:44 PM] |
||
|
|
JSYKES
Senior Cruncher Joined: Apr 28, 2007 Post Count: 200 Status: Offline Project Badges: 
|
The correlation of cores to WU's seems to be consistent (I have 4 WU's and 8 cores) - but so does another trait from the earlier round - I have all 4 WU's crunching at the moment and for the second time (the first was in the earlier round) they have frozen the PC - and yet again nuked another WU or two on restart (one C4C and an HFCC) back to zero - except that two of the betas have restarted from their achieved percentages (32% and 17% approx) whilst the other two have wound back to full zero time restart.....having clearly been originally started before the 17% WU that has restarted from a saved position. There is a potential problem with progress logging and the interval between logs and a bigger problem with restarts - particularly if real world crunchers turn off their machines at the end of workday - or there are further streams of Microsoft security updates causing restart 'waves' worldwide.
----------------------------------------On the I/O side, I have to say that there doesn't seem to be more HD activity than usual and the memory usage appears to be stable at a low 20-25% leaving 9Gb of headroom.....but I can do without the hanging lark - that is a complete pain as is the long time between log saves.....over to you Tech Guys to sort that one out.  |
||
|
|
|