| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Completed Research
Forum: The Clean Energy Project - Phase 2 Forum
Thread: Upcoming server downtime: May 5th, 2014 |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 63
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hello everybody!
Back in August the research computing (RC) team here at Harvard started a major overhaul of the computing and server resources for the entire University. Now that spring is here and we are about to do some renovation, the room where we store the jabbas and the CEP servers needs to be decommissioned. Our friends at RC have very generously offered to move our servers and the storage jabbas into their secure data center in downtown Boston. In the long run this will mean that the CEP servers get a more professional love and attention. This move will be happening on Monday, May 5th. We are currently aiming to have the machines relocated and running on the evening of that day. Worse case something fails to start on the move and we may need to take a little of Tuesday. Since we'll be moving the server machines that process the data being fed from the World Community Grid, we'll need to pause the feed during this move. The IBM team can probably provide a better overview of what this temporary server downtime means from their side of the grid. Thank you in advance for your understanding during this temporary downtime. - Your Harvard CEP team |
|||
|
||||
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
To track the status of the server outage please follow the news article at http://www.worldcommunitygrid.org/about_us/viewNewsArticle.do?articleId=357
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Be warned: If a node has equal or more than 2 x ncpus of cep2 buffered, work fetching from wcg will stop as soon as the agent has accumulated 2x ncpus of completed results that cannot be uploaded. In the example, if a dual threaded node has 4 or more cep2 tasks buffered and the 4th is completed without ability to upload, new work cannot be downloaded. You will get something like "too many results to upload, not sending new work". One solution: Pre-buffer more work than the estimated duration of the outage, -plus- a good number of hours extra as their will very probably be upload cram at harvard once they go online again. Another solution: temporarily set the allowed number of cep2 in the device profiles to less than 2 x ncpus and select other wcg sciences to fill the time, or activate a backup project to keep agents busy.
----------------------------------------[Edit 2 times, last edit by Former Member at May 2, 2014 12:59:11 PM] |
||
|
|
Gatchaman
Cruncher Joined: Feb 29, 2012 Post Count: 49 Status: Offline |
Good luck with the move. And see you on the other side!
----------------------------------------Danm it ...again! Just noticed my post count has reset after changing my user name and my signature thingy has gone too. Double damn it! :-) And....it's back.  "Sadly this project is turning into nonscience......" [Edit 2 times, last edit by Gatchaman at May 4, 2014 9:30:28 PM] |
||
|
|
RichSavarie
Cruncher Canada Joined: Aug 9, 2005 Post Count: 49 Status: Offline Project Badges: 
|
Well then. This explains the lack of upload success I've had today. I guess I'll just ignore this until sometime tomorrow. Thanks for the heads up.
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Any news yet?
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The hands on news is that my first and only cep2 result part _4 fails to upload, cycling from 0 to 100 percent and back. Now on a 1:54 hour back-off counter. Left it alone, going to leave it alone.
|
||
|
|
Gatchaman
Cruncher Joined: Feb 29, 2012 Post Count: 49 Status: Offline |
Actually this comes at a really good time for me as I had runout of CEP2 work and decided to migrate my work pc OS to my first ever ssd. Okay that took me a couple of hours and figuring out how to hide old partitions was fun for a while but I guess your move is a bit more complicated than mine ;-).
----------------------------------------"Sadly this project is turning into nonscience......" |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Work was supposed to be completed by the end of 5/05/14 - it is now almost the start of 5/07/14 - How is the move going & when can we expect to receive new cep2 tasks?
|
||
|
|
jonnieb-uk
Ace Cruncher England Joined: Nov 30, 2011 Post Count: 6105 Status: Offline Project Badges: 
|
From the News article Temporary pause of the Clean Energy Project
----------------------------------------We will update you on the status of the migration by posting updates to this article. It seems the new Communications features and their implementation are still not entirely fit for purpose! Updates to the status of the migration should automitically include an update if completion is delayed beyond the original target. Acccording to cleanenergy's original post the target for complation was Monday evening and yet 12+ hours later and no update has been forthcoming. |
||
|
|
|