| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Official Messages
Forum: News
Thread: 2023-11-24 (Weekend work unit availability) |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 68
|
|
| Author |
|
|
ericinboston
Senior Cruncher Joined: Jan 12, 2010 Post Count: 265 Status: Offline Project Badges: 
|
Can you imagine, that people have better things to spend their weekends on than their fu**ing job, on which they already spend the whole week? In particular if they are not getting payed for working on weekends (and I'm sure WCG staff isn't)? There is no urgency, this is not something anybody's life depends on! Thank you for not reading my comment. And by the way, yeah, when a worldwide project (non-profit, for-profit, volunteer, whatever) goes offline, someone at WCG should have been alerted (automatically from the cloud provider, datacenter, scripts, networking monitors, app monitors, etc.) within minutes and someone at WCG should have AT LEAST communicated to the project volunteers within a few hours that WCG IS AWARE of the problem and maybe even a ballpark ETA to be fixed. But no. There may be no urgency in fixing the problem within X hours, but the lack of communication reflects poorly on WCG and volunteers will simply get frustrated and walk away. I'm EXTREMELY passionate about WCG and the Cancer projects it supports. If WCG can't do the basics with uptime and communication, I'll take my volunteer cycles elsewhere. I really do not want to do this, but maybe this is the 2nd to last straw for me. Either that or set up backup projects, whatever is better suited to prevent such completely unnecessary outbursts of anger in the future. I suggest backup projects. Um, all of WCG is down (implied by others on this thread listing numerous projects that are down as well as WCG posting on Twitter this morning that all of WCG is down). So backup projects would be of no help. I'm not allowed to share my opinion and be angry/unhappy? I/we can only share positive opinions in this forum? Sounds a bit biased if you ask me. If I were swearing and making personal attacks, sure, I would agree that's unproductive, unprofessional, and unnecessary.  [Edit 2 times, last edit by ericinboston at Dec 4, 2023 4:25:07 PM] |
||
|
|
ericinboston
Senior Cruncher Joined: Jan 12, 2010 Post Count: 265 Status: Offline Project Badges:
|
WCG *seems* to be back up. As of 2023-12-04 15:56:16 UTC I am seeing uploads on the Results webpage.
---------------------------------------- |
||
|
|
Link64
Senior Cruncher Joined: Feb 19, 2021 Post Count: 206 Status: Recently Active Project Badges: 
|
And by the way, yeah, when a worldwide project (non-profit, for-profit, volunteer, whatever) goes offline, someone at WCG should have been alerted (automatically from the cloud provider, datacenter, scripts, networking monitors, app monitors, etc.) within minutes and someone at WCG should have AT LEAST communicated to the project volunteers within a few hours that WCG IS AWARE of the problem and maybe even a ballpark ETA to be fixed. I expect this from my bank and I would expect this from other similarly important services, but for sure not from a BOINC project.I'm EXTREMELY passionate about WCG Yes, that's clearly the issue you need to work on and realize, that BOINC projects are not any kind of critical services, which must function 24/7 and pay for staff monitoring the servers and keeping the volunteers updated all the time so that they don't have any nervous breakdowns. And yes, being 24/7 on standby for posting information here is a job, for which someone needs to be paid if they should do it at any time of the day as you expect. No one will or should do that perhaps in the middle of the night or in general in their spare time for free just to please you.If WCG can't do the basics with uptime and communication, I'll take my volunteer cycles elsewhere. I really do not want to do this, but maybe this is the 2nd to last straw for me. Why not just crunch for more than one project? That's what BOINC was made for.Um, all of WCG is down (implied by others on this thread listing numerous projects that are down as well as WCG posting on Twitter this morning that all of WCG is down). So backup projects would be of no help. BOINC is not just WCG, there are many other active projects, some of them have been suggested as backup several times, also for the time when WCG moved from IBM to Krembil. There's also this thread, at the bottom of the first post you have a link to a list of currently active projects. Or simply check BOINCstats or other stats sites for active projects. There are many ways to find BOINC projects. [Edit 3 times, last edit by Link64 at Dec 4, 2023 5:19:26 PM] |
||
|
|
BobbyB
Veteran Cruncher Canada Joined: Apr 25, 2020 Post Count: 638 Status: Offline Project Badges: 
|
Agreed about not having to be on call 24/7. Well at least not for non-critical stuff and if they are then the $$$ should be there or other compensations provided.
I do expect this at NORAD. I'm not allowed to share my opinion and be angry/unhappy? I/we can only share positive opinions in this forum? I have noticed this just a bit but anyway.....I do have backup projects and I did manage to keep my more powerful machines fed. My 4-core laptops are off now and they are only there because I found it better than junking them. I only do human science and/or medical science. Denis and Tn-Grid are both dry. Rosetta came in to the rescue on one and Sidock on the two others. I dislike Sidock for 2 reasons. One is political and the other is that their WUs are all 1 day+ in length. It's a killer and only go to it as a last resort (Добро пожаловать, Влад). When ARP was around, I never let those 4-core laptops near them. Bottom line: I did not lose any time. |
||
|
|
ericinboston
Senior Cruncher Joined: Jan 12, 2010 Post Count: 265 Status: Offline Project Badges:
|
...BOINC projects are not any kind of critical services, which must function 24/7 and pay for staff monitoring the servers and keeping the volunteers updated all the time so that they don't have any nervous breakdowns. And yes, being 24/7 on standby for posting information here is a job, for which someone needs to be paid if they should do it at any time of the day as you expect. No one will or should do that perhaps in the middle of the night or in general in their spare time for free just to please you. 1)I have never stated nor been under the impression that BOINC projects (especially WCG) are a critical service. 2)This is not 1995 or 1985. IT people are not sitting at their desks actively watching the various components/tools/apps of WCG waiting for a failure. This is almost 2024. WCG is (or darn well should be) cloud computing centric...which in turns offers the most basic monitoring tools that would INSTANTLY alert some WCG staffer about any given problem. What I expect (and so do others, even if not voicing themselves here (most likely because they don't even know about the forums)) is BASIC communication regarding a complete outtage like this weekend. My examples of improving this communication have been articulated earlier in this thread. My expectation of such communication is not unreasonable, especially given the scale and geographic reach of WCG. Lack of communication (all kinds of communication) has been a hot topic for many, many years. Now that Krembil owns WCG, it seems nothing has changed. Communication is so, so, so easy to do and is the "face" of any project. [Edit 2 times, last edit by ericinboston at Dec 4, 2023 5:47:21 PM] |
||
|
|
Link64
Senior Cruncher Joined: Feb 19, 2021 Post Count: 206 Status: Recently Active Project Badges:
|
1)I have never stated nor been under the impression that BOINC projects (especially WCG) are a critical service. Well, than what's the issue if the servers are down for a day or two while the WCG staff can enjoy their weekend without being alerted about some server hickups and spending their time on posting in the forums?2)This is not 1995 or 1985. IT people are not sitting at their desks actively watching the various components/tools/apps of WCG waiting for a failure. This is almost 2024. WCG is (or darn well should be) cloud computing centric...which in turns offers the most basic monitoring tools that would INSTANTLY alert some WCG staffer about any given problem. It doesn't matter how WCG is running, people want to be paid if they should do anything outside their working hours and from what I can tell from the previous outages here, no one from WCG team and also no one in the "data center" is working on the weekends or in general outside their working hours. That's not different on many other BOINC projects (and most of them are even worse with communication). The only thing that sucks here is the lack of server status page, but I guess it's hard to bring it back to this highly customized BOINC environment. That page would be telling us all we need to know during the weekend, more information could be than posted on Mondays.Communication is so, so, so easy to do Sure. But it's not free and WCG@Krembil does not seem to have lots of money in genaral considering the hardware issues they had at the beginning and also other information they posted. |
||
|
|
ericinboston
Senior Cruncher Joined: Jan 12, 2010 Post Count: 265 Status: Offline Project Badges:
|
...But it's not free and WCG@Krembil does not seem to have lots of money in genaral considering the hardware issues they had at the beginning and also other information they posted. Look, you support/defend their poor lack of communication and support because of some perceived cost that is unbearable by Krembil (and in the past IBM). This lack of communication and support directly refute Krembil's mission and public website statements: https://www.worldcommunitygrid.org/about_us/article.s?articleId=732 highlights at the very top "We're thrilled to announce that World Community Grid will now be run and supported by Krembil Research Institute, one of the world's leading basic science and clinical research institutions." In the 1st sentence of the above link, they point to https://www.uhn.ca/Research/Research_Institut.../Pages/about_Krembil.aspx which proudly states "Our goal is to be one of the top five research institutes, worldwide. " Krembil can't be touting how great they want to be (or are) while providing terrible service to 100,000+ active volunteers around the world that are passionate about supporting THEIR project (WCG) that, in their own words, "will now be run and supported by Krembil Research Institute" I'm done on this conversation and have no ill toward you or others who may disagree with my opinions. My goal (for 10+ years now) is to passionately dedicate CPU time to Cancer projects on WCG. I hope someone at WCG will read this relatively short thread and make changes. |
||
|
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 3010 Status: Offline Project Badges: 
|
From the passionate coversation on here, let me add a few words from "the other side".
----------------------------------------I work for a very large high street bank in the UK, providing IT support to 3 critical/customer facing systems on a 3 weekly 'on-call' cycle, thus, for the week when I AM on call, I've got my phone with me 24/7, for the whole 168 hours in that week. After doing this for a good number of years, I can honestly say as to how wearying this is - not knowing IF or WHEN the phone will ring, and IF it does, what the issue will be (e.g., if it's a "two miniute issue" or a long-drawn out and complicated failure). This results in a very poor week of sleep, as I've got to have my phone next to me during the night, and woebetide me if I don't answer it if/when it rings. Now, this is the deal I signed up for (I've been working in IT support for over 30 years now), and as I'm supporting CRITICAL systems that my banks customers rely upon, I'm compensated well for doing this. I certanliy would NOT expect anyone to do this for non-critical systems (such as any BOINC project - including WCG), as it does rob me of my social life and also wearies me for performing my 9-5 role. Yes, it would be NICE to have 24/7 cover/communication experts providing us all with a constant 'minute-by-minute/hour-by-hour" update of what the issue is, but the money that this would cost, can (and should, in my eyes), be put elsewhere. If WCG stops working for a few days, it's not life-threatening/people won't be able to pay their rent/buy food etc., (such as would happen if the systems I support, failed).  |
||
|
|
TPCBF
Master Cruncher USA Joined: Jan 2, 2011 Post Count: 2173 Status: Offline Project Badges: 
|
WCG *seems* to be back up. As of 2023-12-04 15:56:16 UTC I am seeing uploads on the Results webpage. Well, it all depends on how one defines "being up".While all WUs have been reported and new WUs (for both MCM1 and OPNG) have been issued, it seem like none of those reported MCM1 WUs have been validated. Haven't looked to see if any of the wing men are MIA still, but the oldest WU right now is from late 11/28, that's at least early AM on Thursday in Toronto and I tried to post about the fact that something seems to be off on Friday. Well before the weekend, so not even the need for anyone to not being able to enjoy their weekend. And the communication issues are going on for pretty much a year and a half now, it took even weeks to acknowledge that there is a problem with those MCM1 WUs not being purged and building up. Which also makes me wonder if even with the new storage that was talked about, the system is running low on resources and is causing the issues from Friday on. We have now one whole day before the whole system is supposed to go down for planned maintenance, let's see how things are looking come Thursday morning... Ralf |
||
|
|
hchc
Veteran Cruncher USA Joined: Aug 15, 2006 Post Count: 865 Status: Offline Project Badges: 
|
@ericinboston:
----------------------------------------Whoa, ~2300 years! I'd be interested in knowing your hardware farm setup. Must be one heck of an electric bill. I'd say @tag me in a thread in the Hardware forum, but this forum software is egregiously obsolete and doesn't support tagging and likely has a ton of security vulnerabilities. I hope to be in a position someday to have hundreds of cores/threads going at it, but I'm lucky to have some quad core machines here and there crunching away.
|
||
|
|
|