Message boards :
Number crunching :
adequately supply or WUs: PLEASE: increase max_wus_in_progress or enable the Computing preference settings
Message board moderation
Author | Message |
---|---|
Send message Joined: 30 May 13 Posts: 18 Credit: 5,655,668 RAC: 34 |
I like Milkyway@home as a project. Among other things, my Dad was a career NASA engineer, and I myself also worked for NASA at Cape Canaveral for a couple of years. Thus, I like to run Mikyway as my primary BOINC project. However, I cannot take the time to ‘babysit’ it day and nite, and make sure that there enough tasks (WUs) to keep my 4 PCs and a laptop adequately supplied with WUs PLEASE: increase max_wus_in_progress from the current setting of 3 or enable the Computing preference settings: “Maintain enough tasks to keep busy for at least ___ days†“… and up to an additional ___ days†L Pallos, PhD LPallos.ie87@GTalumni.Org LLP, PhD, Prof. Engr. I think, therefor I THINK I am. God is Love, Jesus proves it. God is Love ... all (well -- most, anyway) project stats |
Send message Joined: 30 May 13 Posts: 18 Credit: 5,655,668 RAC: 34 |
I do not see a way to edit a theme name once posted. The theme was meant to be: adequate supply of WUs: PLEASE: increase max_wus_in_progress or enable the Computing preference settings Thank you LLP, PhD, Prof. Engr. I think, therefor I THINK I am. God is Love, Jesus proves it. God is Love ... all (well -- most, anyway) project stats |
Send message Joined: 8 Feb 08 Posts: 261 Credit: 104,050,322 RAC: 0 |
Milkyway needs short turnaround times because the new WUs are generated based on the results coming in. If you check this board, you will find several discussions about it. |
Send message Joined: 19 Jul 10 Posts: 627 Credit: 19,303,016 RAC: 1,302 |
You can use the backup project feature of BOINC for the cases, when Milkyway is out of work (which didn't happen often recently). On a project, which you would like to use as backup, simply set the resource share to 0 and enable work fetch for that project, if your BOINC client can't get enough work from Milkyway, than it will ask for work from the backup project. |
Send message Joined: 23 Sep 12 Posts: 159 Credit: 16,977,106 RAC: 0 |
I am looking into the impact of more WUs in progress on the efficiency of the server algorithm. But more specifically are others having issues with the WUs not be generated often enough or requesting new WUs. Jeff |
Send message Joined: 23 Sep 12 Posts: 159 Credit: 16,977,106 RAC: 0 |
I would also check out this thread and the links in the thread to previous discussions on the issue. http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=1818#40607 Jeff |
Send message Joined: 30 May 13 Posts: 18 Credit: 5,655,668 RAC: 34 |
Thanks for looking into increasing the number of WUs that can be 'in progress' For the record: apparently a few years ago the limit was 6 per CPU: Posted: 23 Jun 2010, 19:41:24 UTC It is 6 tasks per cpu core. I am not sure if this is literally correct, or if the limit even then was really 3 per CPUC core X2 with threading. I certainly am not asking for "3 days of work" However, only allowing 3 WU per cpu when the WUs are only ~55 min long is ONLY allowing about 2 HOURS, 45 MINUTES of work. Additionally, Would MW please take a good look at the scheduling&WU-requesting algorithm -- please? Why is that MW will repeatedly request more WU minute by minute when the limit had just recently been filled, and then later NOT re-schedule another request for a couple of hours Lászlo Pallos, PhD Prof. Engr. LLP, PhD, Prof. Engr. I think, therefor I THINK I am. God is Love, Jesus proves it. God is Love ... all (well -- most, anyway) project stats |
Send message Joined: 13 Mar 08 Posts: 804 Credit: 26,380,161 RAC: 0 |
Laszlo as you report WU's they should get replaced with new ones. |
Send message Joined: 30 May 13 Posts: 18 Credit: 5,655,668 RAC: 34 |
they should get replaced Yes, they should. However, sometimes WM's WU fetch algorithm takes a loooong break, and then the replacements arrive after my machine's been idle for a while. Does MW really need to limit WU's in progress to < 2 hr 45 min?? So, what's the big deal about allowing a 5 or 6 hour 'inventory' of WU's?? This should still give turnarounds on tasks with total adequacy for MW's 'real time' task generation. I repeat, WM is my #1 BOINC project, hands down. But, until this 'stinginess (?)' on WU's is resolved, I have switched to my #2 project. Hope to 'be back soon'... LLP, PhD, Prof. Engr. I think, therefor I THINK I am. God is Love, Jesus proves it. God is Love ... all (well -- most, anyway) project stats |
Send message Joined: 22 Jan 11 Posts: 375 Credit: 64,707,164 RAC: 64 |
I would also check out this thread and the links in the thread to previous discussions on the issue. Lol, mod who didn't link, now linkified :) [url]http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=1818#40607 [/url] Wth?? I can't get the link to work..... Test http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=3465#61494 Well that's just baffling! Why is 1 clickable & the other not???! Team AnandTech - SETI@H, DPAD, F@H, MW@H, A@H, LHC, POGS, R@H, Einstein@H, DHEP, WCG Main rig - Ryzen 5 3600, MSI B450 G.Pro C. AC, RTX 3060Ti 8GB, 32GB DDR4 3200, Win 10 64bit 2nd rig - i7 4930k @4.1 GHz, HD 7870 XT 3GB(DS), 16GB DDR3 1866, Win7 |
Send message Joined: 18 Jul 09 Posts: 300 Credit: 303,565,482 RAC: 0 |
I would also check out this thread and the links in the thread to previous discussions on the issue. |
Send message Joined: 15 Sep 12 Posts: 20 Credit: 105,341,548 RAC: 0 |
I am looking into the impact of more WUs in progress on the efficiency of the server algorithm. But more specifically are others having issues with the WUs not be generated often enough or requesting new WUs. Jeff and yet somehow I'm managing close to 1 million a day on the current challenge on Boincstats and not running out of units ? and never had an issue getting work to keep a minimum of 7 computers crunching ? no problem getting units to crunch here. |
Send message Joined: 8 May 09 Posts: 3339 Credit: 524,010,781 RAC: 0 |
I am looking into the impact of more WUs in progress on the efficiency of the server algorithm. But more specifically are others having issues with the WUs not be generated often enough or requesting new WUs. I don't have a problem either but live in the US, the person asking lives in El Salvador, I wonder if they are having intermittent internet problems, them or the provider? Here in the US I also have constant internet that is always on for my flat fee, I have seen where others have to pay based on the usage and time of day, I wonder if that is the underlying problem? |
Send message Joined: 22 Jan 11 Posts: 375 Credit: 64,707,164 RAC: 64 |
In the UK here & no problem with having enough WUs, although I have noticed in todays log a regular problem of not getting a task when asked. swiftmallard Lol, still don't know my 1st link didn't work, their are no typos! Let's see if it works this time! http://milkyway.cs.rpi.edu/milkyway/forum_thread.php?id=1818#40607 Yep! weird, lol. Team AnandTech - SETI@H, DPAD, F@H, MW@H, A@H, LHC, POGS, R@H, Einstein@H, DHEP, WCG Main rig - Ryzen 5 3600, MSI B450 G.Pro C. AC, RTX 3060Ti 8GB, 32GB DDR4 3200, Win 10 64bit 2nd rig - i7 4930k @4.1 GHz, HD 7870 XT 3GB(DS), 16GB DDR3 1866, Win7 |
Send message Joined: 7 Nov 13 Posts: 4 Credit: 2,115,550 RAC: 15 |
Other projects have work-fetch algorithms which are triggered whenever a current WU is completed. This is not only an easy and a natural logic rule, it also helps to optimize the work flow. Please modify the Milky Way work-fetch algorithm to include a component of requesting new work at any time when a running WU completes. Thank you. LLP, PhD, Prof. Engr. God is Love |
Send message Joined: 8 May 09 Posts: 3339 Credit: 524,010,781 RAC: 0 |
Other projects have work-fetch algorithms which are triggered whenever a current WU is completed. This is not only an easy and a natural logic rule, it also helps to optimize the work flow. If you change your workunit cache to zero point one in both the min and additional boxes you will have done the same thing on your own pc's. You will have one unit per allowed cpu core/gpu, and when it is done you will get a new unit. |
Send message Joined: 19 Jul 10 Posts: 627 Credit: 19,303,016 RAC: 1,302 |
Other projects have work-fetch algorithms which are triggered whenever a current WU is completed. This is not only an easy and a natural logic rule, it also helps to optimize the work flow. This is not Milkyway thing, this is BOINC thing, after all it's your BOINC client that requests work and not the Milkyway servers. Set your cache large enough and BOINC should request new WUs as soon as it finishes one (since finishing a WU should reset any backoffs). Make sure to set the cache properly for the BOINC version you use, v7 needs different settings than v6 and older clients (your computers are hidden, so no idea which you use). Eventually post your settings here if you are not sure. |
Send message Joined: 22 Jan 11 Posts: 375 Credit: 64,707,164 RAC: 64 |
Are you sure about that? I have a feeling MW uses some kind of an over ride on cache settings........ Team AnandTech - SETI@H, DPAD, F@H, MW@H, A@H, LHC, POGS, R@H, Einstein@H, DHEP, WCG Main rig - Ryzen 5 3600, MSI B450 G.Pro C. AC, RTX 3060Ti 8GB, 32GB DDR4 3200, Win 10 64bit 2nd rig - i7 4930k @4.1 GHz, HD 7870 XT 3GB(DS), 16GB DDR3 1866, Win7 |
Send message Joined: 30 May 13 Posts: 18 Credit: 5,655,668 RAC: 34 |
Thanks for your response posts. I am inclined to agree with [TA]Assimilator1. I have pretty much set milkyway preferences the SAME as I have for all the other projects I run, and for the other projects, the finishing of a WU automatically sends a request for more WUs I could see your point, Link, about the BOINC manager. However, again, why do other projects send a request on a WU completion and MW does not? Another counterpoint to this being a BOINC manager issue: I set a request for additional work on "Computing preferences" (Maintain enough tasks to keep busy for at least ... ... and up to an additional 0.5 days ) This DOES get BOINC to request more work, but 1) not more frequently and 2) the size of the request is overridden by MW. LLP, PhD, Prof. Engr. I think, therefor I THINK I am. God is Love, Jesus proves it. God is Love ... all (well -- most, anyway) project stats |
Send message Joined: 30 May 13 Posts: 18 Credit: 5,655,668 RAC: 34 |
by 'cache' do you mean the Preference settings: Maintain enough tasks to keep busy ... LLP, PhD, Prof. Engr. I think, therefor I THINK I am. God is Love, Jesus proves it. God is Love ... all (well -- most, anyway) project stats |
©2024 Astroinformatics Group