Message boards : Projects : Anything and Everything to do with (WCG) World Community Grid
Message board moderation
Previous · 1 · 2 · 3 · 4 · 5 . . . 14 · Next
Author | Message |
---|---|
Send message Joined: 30 Mar 20 Posts: 392 |
Some people now seems to have been able to at least upload their finished work. Not so lucky here though, at least not yet. https://www.worldcommunitygrid.org/forums/wcg/viewpostinthread?post=683363 |
Send message Joined: 3 Mar 23 Posts: 14 |
Threaten me with nukes? You got me! LOL (= |
Send message Joined: 30 Mar 20 Posts: 392 |
Apparently, it was a short period today when at least the upload server was running, since some people managed to upload some of their finished tasks. Or their tasks simply went into the wide blue yonder, never to be seen again. But: Last update host XML 2023-02-28 13:11:22 UTC (22 days 20:59:04 old) Last update user XML 2023-03-01 01:21:01 UTC (22 days 08:49:25 old) Last update team XML 2023-03-01 01:21:01 UTC (22 days 08:49:25 old) |
Send message Joined: 10 May 07 Posts: 1423 |
WuProp website & project servers UNREACHABLE AGAIN !!!!! Been down about 45 minutes |
Send message Joined: 8 Mar 23 Posts: 10 |
there's a good chance these technical issues will result in loss of data. WUs completed months or years ago could be lost - and i bet they'll never tell us. |
Send message Joined: 30 Mar 20 Posts: 392 |
there's a good chance these technical issues will result in loss of data. WUs completed months or years ago could be lost - and i bet they'll never tell us.Yes, I have also thought about that. That's why I wrote about "the wide blue yonder". And I agree that it's likely that they will never say anything about that. |
Send message Joined: 25 May 09 Posts: 1295 |
While they may not know about an individual user loosing data the way BOINC works on the server make sure that the results for a task sent out but never returned are sent out to another user. This may be a bit hard on the individual user, but the science data is pretty well protected. |
Send message Joined: 30 Mar 20 Posts: 392 |
While they may not know about an individual user loosing data the way BOINC works on the server make sure that the results for a task sent out but never returned are sent out to another user. This may be a bit hard on the individual user, but the science data is pretty well protected.A bit hard yes. I have 10 tasks ready and waiting to be uploaded. If i lose those, my life will be over, and I might as well off myself :-) |
Send message Joined: 7 Apr 13 Posts: 64 |
on a quick update, finally, /science filesystem is on the move to the new storage from the recovery storage unit. As of last night, after 3 hours, the new storage /science filesystem shows 1.4TB used. Assuming such average rate of file transfer, it will take about 74 hours. Hopefully, we will be able to restart BOINC from the new storage and finally put the failure behind us. We will keep you posted. That's an update a few hours ago from someone who appears to be on Krembil / Jurisica team. Here's another post looking ahead - doesn't make much sense and does not sound good... as for the help - logistic is tricky considering we run from a different data centre - and of course we cannot give access to a broad group - but once we can at lest walk again, there are things we plan on our side, and other with the broader community. Briefly - we need to simplify the backend - at the moment, we often run into multi points of failure, instead of robustness. But - once we will be in such a position - we want to run hackathons - this can substantially help with optimizing code we run on the grid, and bring new projects. So far, nVidia is interested to discuss this further - as our plan is to bring more GPU projects. But - of course the backend has to be upgraded before that - as peak performance during GPU stress test in 2021 was around 16PFLOPS. |
Send message Joined: 30 Mar 20 Posts: 392 |
Igor Jurisica is the boss of the Jurisica lab, and not just "someone who appears to be on Krembil / Jurisica team." https://www.cs.toronto.edu/~juris/jlab/members.html https://www.cs.toronto.edu/~juris/jlab/contact.html |
Send message Joined: 28 Jun 10 Posts: 2637 |
Getting close to a full day without a post or grumble in the thread! |
Send message Joined: 10 May 07 Posts: 1423 |
Getting close to a full day without a post or grumble in the thread! OK Dave /Grumble activated/ Get off my lawn! /Grumble deactivated/ |
Send message Joined: 10 May 07 Posts: 1423 |
Found this while browsing the WCG forums today>>> RE: Recovery Update and Donations... |
Send message Joined: 7 Apr 13 Posts: 64 |
C'mon folks we are slipping - there were zero complaints about WCG being down yesterday. This is unacceptable especially since they don't even work on weekends - we must do better. Here's the first complaint today, can someone sign up for this afternoon? Ni! 😆 |
Send message Joined: 30 Mar 20 Posts: 392 |
Last update host XML 2023-02-28 13:11:22 UTC (27 days 00:45:54 old) Last update user XML 2023-03-01 01:21:01 UTC (26 days 12:36:15 old) Last update team XML 2023-03-01 01:21:01 UTC (26 days 12:36:15 old) No further comments needed. |
Send message Joined: 10 May 07 Posts: 1423 |
Okay, I'll start with: It's Monday almost mid-day in the great white north and the silence from Krembil/WCG is nearly deafening over the crickets chirping. |
Send message Joined: 30 Mar 20 Posts: 392 |
Well, I seriously doubt that this day will be the day WCG restarts. Despite the Terabyte calculations Igor Jurisica posted on Friday March 24. https://www.worldcommunitygrid.org/forums/wcg/viewthread_thread,44980_offset,140#683387 |
Send message Joined: 7 Apr 13 Posts: 64 |
...it will take about 74 hours. Sounds good, but don't assume that is around the clock crunching. Spread over 8 hour days and no weekends or a standard 40 hour week, that could be 2 more weeks. So maybe 4/10? |
Send message Joined: 30 Mar 20 Posts: 392 |
It's not about any crunching, it's about moving/copying /science filesystem, to the new storage from the recovery storage unit. And that should be done without any human intervention as far as I understood it, from other posts by Igor....it will take about 74 hours. But of course, I would not be surprised at all, if the whole procedure crashed and burned, some time during the weekend. |
Send message Joined: 28 Jun 10 Posts: 2637 |
Having been a little bit on the inside in my role as a moderator for CPDN when they had hardware problems with running out of space from a new model type producing data a lot faster than it could be moved. I also would not be surprised if things take longer than predicted. |
Copyright © 2024 University of California.
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License,
Version 1.2 or any later version published by the Free Software Foundation.