| 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: Research Log: Updates from the Harvard Team |
| No member browsing this thread |
|
Thread Status: Active Thread Type: Sticky Thread Total posts in this thread: 1840
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Please note that the WCG team currently performs a scheduled outage on their servers to upgrade to the latest IBM General Parallel File System software. During this time, the World Community Grid (BOINC) clients will not be able to upload or download files nor will they be able to report completed results or fetch new work. We’ll keep you posted once everything is back to normal.
http://www.worldcommunitygrid.org/forums/wcg/viewthread_thread,32741#372084 |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
There was some renewed interest in the details of the CEP hardware setup, so here is a summary of our current infrastructure: The Harvard side of CEP uses two Dell servers with DNS round-robin to receive the data from the grid. The servers sit on Ms Piggy, a fast 48TB storage array consisting of 2 Dell MD1200s. After some initial postprocessing, the data is moved to the slower, high-capacity storage arrays called Jabba. The Jabba arrays are based on a design by Backblaze:
http://blog.backblaze.com/category/tech-news/ Jabba01 followed the v1.0 design, which we had to update a good bit: http://blog.backblaze.com/2009/09/01/petabyte...uild-cheap-cloud-storage/ We are currently retrofitting our original 2TB Seagate setup with 3TB Hitachi drives. The nominal capacity is 105TB right now. Jabba02 and 03 use the v2.0 design and we bought prefabricated boxes from Protocase. The 2.0 version uses 3TB Hitachi drives and turned out to be a good bit more stable than v1.0. Both are set up as 135TB arrays. http://blog.backblaze.com/2011/07/20/petabyte...-0revealing-more-secrets/ So, once jabba03 is fully operational, we will have a nominal capacity of 48+105+135+135=423TB, however some of this used up by the RAID6. The data is also backed up on tape with the new tape robot at FAS-RC. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
It appears that yesterday's WVG server upgrade went smoothly and we seem to be back to normal now. Good work by the IBM-WCG Team (as always)!
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
It appears that yesterday's WVG server upgrade went smoothly and we seem to be back to normal now. Good work by the IBM-WCG Team (as always)! Not to steal the Clean Energy team's favorite weekend intro, but seeing as this is one of my all-time favorite cartoons and the IBM-WCG team just slid an upgrade in nice and slick: https://xkcd.com/705/Is one of the few things that make me laugh every time. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The 3 TB drives are state of the art today, high cap drives, in two years you can sell them on e bay for 10 bucks a pop heh.
Just out of curiosity, are your RAIDS smart enabled, so you can get a clue that something is going bad before it actually does? Although the end result is still the same, swap the drive and let it clone... Aaron |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hi ibsteve2u,
yes, that's indeed an awesome one. Have to point this out to our friends at FAS-RC ;). Best wishes from Your Harvard CEP team |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hi Ascholten,
Yes, we set up our RAID arrays to be SMART enabled. We get daily emails on the status of every one of the disks (we set up daily short and weekly long tests). We also get a fair warning when a disk is in trouble. Have a great weekend from Your Harvard CEP team |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Hi everybody,
----------------------------------------Our call for suggestions to name our new in-house cluster was a big success. We received a ton of great suggestions on the WCG forum, facebook, and twitter. Thanks so much for all your input! In the first selection round, all the candidates were rated by an internal jury. Some suggestions did not quite make the cut for formal, technical, or practical reasons. The best 24 candidates were chosen for the final round and now YOU have the chance to select the ultimate winner of this little competition. You can vote following the link below! May the best name win  http://www.surveymonkey.com/s/2NPHLYS [Edit 1 times, last edit by Former Member at Apr 9, 2012 3:44:13 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
It's time for some broken records again: CEP has now collected over 9,000 years of computing time and produced 11 million results. This is an awesome achievement, everybody! Thanks so much for all your hard work!
----------------------------------------http://www.worldcommunitygrid.org/stat/viewProject.do?projectShortName=cep2 [Edit 1 times, last edit by Former Member at Apr 10, 2012 9:02:41 PM] |
||
|
|
|