| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: Smash Childhood Cancer
Thread: Is there any way of finding your wingman’s host-Id? |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 65
|
|
| Author |
|
|
Grumpy Swede
Master Cruncher Svíþjóð Joined: Apr 10, 2020 Post Count: 2494 Status: Offline Project Badges: 
|
Tiger Lily Totally agree with you Mike.There is another aspect to this cluster problem. It would seem to be Linux and as a consequence, all those re-sends have to go to Linux machines. Those of us with Windows or Android machines do not get a chance for the re-sends. Mike And when looking at BoincStats Hosts Overview for WCG, there are 76.05% Windows hosts, 12.51% Linux hosts, and 11.44% Other. That of course means that Windows hosts are being severely starved by that cluster. That cluster, or whatever it is, needs to be dealt with, pronto. Look at the pie charts at the lower part of this BoincStats page, and the pie chart for: Hosts by Operating System: https://www.boincstats.com/stats/15/project/detail/host [Edit 1 times, last edit by Grumpy Swede at Sep 27, 2023 11:01:49 PM] |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7844 Status: Offline Project Badges: 
|
Tiger Lily There is another aspect to this cluster problem. It would seem to be Linux and as a consequence, all those re-sends have to go to Linux machines. Those of us with Windows or Android machines do not get a chance for the re-sends. Mike Absolutely correct. My Windows system has been out of work pretty much all day. My Linux systems are crunching away on full queues with all workunits being resends. Cheers
Sgt. Joe
*Minnesota Crunchers* |
||
|
|
TigerLily
Senior Cruncher Joined: May 26, 2023 Post Count: 280 Status: Offline Project Badges: 
|
Tiger Lily Hi Mike,There is another aspect to this cluster problem. It would seem to be Linux and as a consequence, all those re-sends have to go to Linux machines. Those of us with Windows or Android machines do not get a chance for the re-sends. Mike Thanks for this additional information. I have passed this on to the rest of the team. |
||
|
|
supdood
Senior Cruncher USA Joined: Aug 6, 2015 Post Count: 333 Status: Offline Project Badges: 
|
Even some of us with Linux machines aren't getting anything. I have one that gets SCC tasks periodically and another that says there are no SCC tasks available. Not sure how SCC can have tasks to download for one computer and none for another running the same OS distro and version requesting work at the same time. Sure would be nice if this cluster either starts doing real work or goes away.
---------------------------------------- |
||
|
|
Sgt.Joe
Ace Cruncher USA Joined: Jul 4, 2006 Post Count: 7844 Status: Offline Project Badges:
|
My Linux queues are still full with almost every unit a resend.. The good news is my pending validation count is down from over 7000 to under 4000.
----------------------------------------Cheers
Sgt. Joe
----------------------------------------*Minnesota Crunchers* [Edit 1 times, last edit by Sgt.Joe at Sep 29, 2023 4:41:53 PM] |
||
|
|
alanb1951
Veteran Cruncher Joined: Jan 20, 2006 Post Count: 1317 Status: Offline Project Badges: 
|
Even some of us with Linux machines aren't getting anything. I have one that gets SCC tasks periodically and another that says there are no SCC tasks available. Not sure how SCC can have tasks to download for one computer and none for another running the same OS distro and version requesting work at the same time. Sure would be nice if this cluster either starts doing real work or goes away. Yup - I've seen this as well! (All Linux systems here...); Interestingly, it doesn't seem to happen with MCM1!My two older (and slower) systems have been keeping topped up to the limits I have set in my WCG profiles, but my two newer systems ran out completely and seemed to struggle to get work most of the time, sometimes getting "No tasks are available" and other times "Tasks are committed to other platforms" (whilst my slower machines were simultaneously getting SCC1 work![*1]) The "slow" systems tend to ask for new work often enough that only one or two replacement tasks are ever needed to top things up at any time, whereas the "fast" systems could finish off several more tasks between top-ups... So I wondered whether it might be something to do with how much work one asks for at once, but cutting down my buffer sizes didn't make any difference (it went straight from "other platforms" to "not needed" when it got low enough!) and I'm not going to mess with my profiles, given the issues reported elsewhere regarding profile changes at present :-) For what it's worth, both the slower machines are Intel whilst the faster ones are AMD Ryzens; that shouldn't make a difference, but there's some strange stuff in the platform-matching code in the [standard] BOINC scheduler, so who knows?... All that said, in my experience it only seems to be an issue when there are [lots of] retries floating around, and my systems run other work as well as SCC1, so for me it's a puzzle rather than a crisis :-) -- users who want to focus on SCC1 will see it differently, and rightly so! Cheers - Al. [*1] I do realize that "simultaneously" is highly unlikely to mean that the machines are asking for work during the same window of availability -- tasks are fed in relatively small batches, but very often. However, it's still strange behaviour! [Edit 1 times, last edit by alanb1951 at Sep 29, 2023 6:10:28 PM] |
||
|
|
Jake1402
Senior Cruncher USA Joined: Dec 30, 2005 Post Count: 181 Status: Offline Project Badges: 
|
There is no newer version for linux - 7.4.22 is still the stable, recommended version. I have 2 Linux machines and they are running 7.16.6. Got it from the BOINC website. I checked Software Manager and it also contains 7.16.6. I sure wish the Linux version got updated as often as Windoze.
Join the Chicago-IL-USA team!
----------------------------------------2 AMD FX 8320/AMD R9 270X/Win 10 2 AMD FX 8320/AMD RX 560/Linux Mint 20.3 (both computers DOA) Intel Pentium G240/Win 10 [Edit 1 times, last edit by Jake1402 at Sep 29, 2023 6:23:20 PM] |
||
|
|
thunder7
Senior Cruncher Netherlands Joined: Mar 6, 2013 Post Count: 238 Status: Offline Project Badges:
|
You mean here:
https://boinc.berkeley.edu/download_all.php where I read These versions may not be current. 7.16.6 Development version (MAY BE UNSTABLE - USE ONLY FOR TESTING) 7.4.22 Recommended version where the first line makes everything suspect. |
||
|
|
adriverhoef
Master Cruncher The Netherlands Joined: Apr 3, 2009 Post Count: 2346 Status: Recently Active Project Badges:
|
Al wrote:
(All Linux systems here...); Interestingly, it doesn't seem to happen with MCM1! My two older (and slower) systems have been keeping topped up to the limits I have set in my WCG profiles, but my two newer systems ran out completely and seemed to struggle to get work most of the time (..) For what it's worth, both the slower machines are Intel whilst the faster ones are AMD Ryzens Exactly what I'm seeing here! With these plethora of retries for certain SCC1-tasks being distributed, my (slower) Intel machines have no problem keeping up, while my (faster) AMD Ryzens are starving most of the time when it comes to SCC1. Adri |
||
|
|
hchc
Veteran Cruncher USA Joined: Aug 15, 2006 Post Count: 865 Status: Offline Project Badges:
|
I'd like to slap the Linux 5.15 person with a big sloppy stinky fish! I came to this forum, and sure enough, it's affected the whole SCC project. I'm not sure how whoever the owner of that machine is got to "reliable" status, but I'm sure that machine is no longer considered reliable.
----------------------------------------Must be many thousands gobbled up and then neglected. My Windows boxes can't get SCC so I had to re-enable MCM1 for them. Really hope it's an innocent mistake and the owner of the devices is reading this thread and embarrassed.
|
||
|
|
|