| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Active Research
Forum: OpenPandemics - COVID-19 Project
Thread: Wingman ? |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 17
|
|
| Author |
|
|
adriverhoef
Master Cruncher The Netherlands Joined: Apr 3, 2009 Post Count: 2346 Status: Offline Project Badges: 
|
They do seem to be all at the same machine. Yes, they are (probably) the same device: OPN1_0090359_00491_0 OPN1_0090359_00922_0 OPN1_0090359_00490_0 Its hostname is "n5.h.asgw.ru" in all three cases and belongs to the domain asgw.ru which resolves to asgreywolf.ru. User ASGreyWolf is a registered member since 06-04-2021 and hasn't returned any result up to this point. There is a name associated with that website asgreywolf.ru: Айдар Фаттахов. That is a quite common name, see vk.com, therefore there are three hints: - https://vk.com/asschool - github.com/AsGreyWolf - according to this link, Aydar lives in Kumertau (a town in the Republic of Bashkortostan), Russia. - https://habrahabr.ru/users/asgreywolf/ The fourth hint, geektimes.ru/users/asgreywolf/, is the same as the third and resolves to https://habr.com/en/users/asgreywolf/. There, this user belongs to a lot of hubs, one of which is: Development for Raspberry Pi … Now, note the osVersion for the device that lies under the magnifying glass: "Arch Linux ARM [5.10.82-2-rpi-ARCH|libc 2.32 (GNU libc)]". Probably a Raspberry Pi. Coincidence? |
||
|
|
sam6861
Advanced Cruncher Joined: Mar 31, 2020 Post Count: 107 Status: Offline Project Badges: 
|
Work unit _1 task sometimes sent first, then _0, can check any MCM1 and ARP1 workunit results to see some work units have task _1 sent first then _0, and this should normally be up to a few minute apart. Being more then a few week apart may have been stuck with "waiting to be sent" for a long time which can possibly a server problem maybe.
On World Community Grid, looks like each work unit are limited to all tasks having the same CPU (x86 x86_64, ARM) and same platform (Windows, Liinux, Android). My guess is, there could be too little number of Linux ARM devices to quickly send both Linux ARM tasks. ... is a registered member since 06-04-2021 and hasn't returned any result up to this point. Probably wrong user member. User members with device installation 0 have no computers connected to this user. World Community Grid web appears to not show the real user member from a work unit, so I don't know the real user member. |
||
|
|
Umlauf
Advanced Cruncher Joined: Mar 20, 2020 Post Count: 52 Status: Offline Project Badges:
|
Why is changing the send time of _0. I think there is something wrong.
|
||
|
|
sam6861
Advanced Cruncher Joined: Mar 31, 2020 Post Count: 107 Status: Offline Project Badges:
|
I wonder if resend lost tasks change the sent time and due time? If yes, this could possibly be a bug or problem at server side.
----------------------------------------Therer is some Linux ARM users saying they can't get tasks working, with log that shows resend lost tasks and discard due to some version mismatch, possibly repeated resend same tasks again and again. https://www.worldcommunitygrid.org/forums/wcg/viewthread_thread,43977 This is my guess: The repeated client request tasks, server resend lost tasks changes sent time and due time, and client just discarding tasks from a problem and requesting more tasks, in a loop, can cause tasks to get stuck with never ending due time, until computer goes offline. I am not sure if this is what is really happening. [Edit 1 times, last edit by sam6861 at Jan 6, 2022 5:46:33 AM] |
||
|
|
adriverhoef
Master Cruncher The Netherlands Joined: Apr 3, 2009 Post Count: 2346 Status: Offline Project Badges:
|
Being more then a few week apart may have been stuck with "waiting to be sent" for a long time which can possibly a server problem maybe. If I understand you correctly, that's not the problem here. The stuck tasks here keep on being sent to that same Arch Linux ARM device and they keep on being shown as "In Progress". ... is a registered member since 06-04-2021 and hasn't returned any result up to this point. Probably wrong user member. User members with device installation 0 have no computers connected to this user.Could it be that a device will be registered and counted only if it has returned any results? I have 640 registered devices (with a number of returned results) and 25 unregistered devices. The unregistered ones aren't counted, they haven't returned any results. About every day a new task is being (re)sent to that same Arch Linux ARM device: 10:42 01-01-2022: Result name OS type Status Sent time Due / Return time CPUtime/Elapsed Claimed/G 15:52 03-01-2022: Result name OS type Status Sent time Due / Return time CPUtime/Elapsed Claimed/G 14:57 04-01-2022: Result name OS type Status Sent time Due / Return time CPUtime/Elapsed Claimed/G 09:27 05-01-2022: Result name OS type Status Sent time Due / Return time CPUtime/Elapsed Claimed/G 08:59 06-01-2022: Result name OS type Status Sent time Due / Return time CPUtime/Elapsed Claimed/G 12:01 07-01-2022: Result name OS type Status Sent time Due / Return time CPUtime/Elapsed Claimed/G The stuck units make me think of Groundhog Day.  It's a pity that they still pronounce déjà vu as déjà vous.  [Edit 2 times, last edit by adriverhoef at Jan 7, 2022 12:05:38 PM] |
||
|
|
alanb1951
Veteran Cruncher Joined: Jan 20, 2006 Post Count: 1317 Status: Offline Project Badges: 
|
A postscript to this little oddity...
I happened to look in on the two original tasks late on 12th January, and discovered that they have finally moved on... Something (presumably a manual intervention) got the duff ARM tasks flagged as No Reply at 18:30 UTC on the 10th, and the initial wing-man got flagged Invalid as a side-effect when the retries started to go out and return on the 10th. In one case, the retries went to non-ARM Linux, in the other case they went to Windows! It would be interesting to know what, if anything, was done to sort this out -- did it need technician intervention??? -- but I suspect we'll never know! Cheers - Al. P.S. Adri - any more news on the other tasks you mentioned later in the thread? (Without access to workunit or result IDs, I can't find the tasks!) |
||
|
|
adriverhoef
Master Cruncher The Netherlands Joined: Apr 3, 2009 Post Count: 2346 Status: Offline Project Badges:
|
P.S. Adri - any more news on the other tasks you mentioned later in the thread? (Without access to workunit or result IDs, I can't find the tasks!) There was one other identified task, Al, and this one, for a change, turned Valid for the initial wingman: Result name OS type Status Sent time Due / Return time CPUtime/Elapsed Claimed/Grant Glad that you spotted this history in time!  [Edit 1 times, last edit by adriverhoef at Jan 13, 2022 1:31:42 AM] |
||
|
|
|