| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
 |
World Community Grid Forums
Category: Support
Forum: Website Support
Thread: New Server Farm [Resolved] |
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 13
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I just read the initial notice of the change over, but there are several unanswered questions. I might suggest editing the original announcement, adding anything missed (if any).
----------------------------------------0. The announcement states the return time is being managed. Does this mean that all In Progress WU are going to be dried up before the changeover, or just that that you want to maximize the shift of jobs from In Progress to Valid? 1. Can we keep crunching while the change is happening? If so, how many days of downtime are planned? 2. Does the change over require any Beta work? 3. Does the new server farm facilitate the return of DDDT or is the return not a function of the new servers? 4. Does the newly announced bandwidth improvement have any impact on AC@H and specifically the minimum computer requirements? 5. For a user who is not behind a firewall, will the new servers simply appear one day so the computers simply pick up where they left off, uploading work done during the downtime? 6. If WU are being crunched during downtime, will the schedule date be reset to reduce frequency of Too Late once the new servers come up? [Edit 1 times, last edit by Former Member at Jan 1, 2009 5:31:10 PM] |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
0. I think the latter.
1. Yes. As few as possible. Optimistically, less than a day - but I expect unforeseen problems. I'm just an old pessimist. 2. No. 3. No connection. 4. No. The limitation was always client bandwidth. 5. I hope so. 6. I don't think that can be done, but if "Too Late"s occur, I expect the techs to be generous with credit. These answers aren't definitive, but are based on what the techs have told the CAs so far. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
Thanks
|
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
So should we up to our cache of WUs during the transition?
I don't want to run out of work if the server will be offline for a long period of time. I have around .5 days cache on most of my machines at the moment. Will this be sufficient or would 1-2 days cache be recommended? |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
The problem is this: WCG want to reduce the number of work units "in the wild", to reduce the amount of data that needs to be transferred. It is much easier for them simply to load new batches on the new servers after the transition is complete.
That directly conflicts with the natural desire of members to have enough work in their queue for a potentially long outage. What to do? Well, I trust the deadline changes will allow the techs to complete as many in-progress batches as possible. On balance, I think a 2 day buffer would be prudent. I expect the techs will be over-scheduling a few batches in the next few days, as well. |
||
|
|
knreed
Former World Community Grid Tech Joined: Nov 8, 2004 Post Count: 4504 Status: Offline Project Badges: 
|
2 days is fine. If all goes well that should be plenty of time.
|
||
|
|
I need a bath
Senior Cruncher USA Joined: Apr 12, 2007 Post Count: 347 Status: Offline Project Badges: 
|
So should we up to our cache of WUs during the transition? I don't want to run out of work if the server will be offline for a long period of time. I have around .5 days cache on most of my machines at the moment. Will this be sufficient or would 1-2 days cache be recommended? I'm doing the opposite. I will wait to get new tasks until after the transition. In the meantime I will crunch other BOINC projects. Me  |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I'm doing the opposite. I will wait to get new tasks until after the transition. In the meantime I will crunch other BOINC projects. Which is exactly what the Tech would like, in order to reduce the number of In progress jobs. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
knreed wrote it would be fine to buffer 2 days of work to bridge the down time next week. If things go to plan, starting January 8. No preference was expressed. So I'll buffer 2 days from Colorado and return the results to Toronto. During the downtime I'll take the client into Network activity suspend and visit occasionally to see if the front page message changes to go back online.
|
||
|
|
Diana G.
Master Cruncher Joined: Apr 6, 2005 Post Count: 3003 Status: Offline Project Badges: 
|
knreed wrote it would be fine to buffer 2 days of work to bridge the down time next week. If things go to plan, starting January 8. No preference was expressed. So I'll buffer 2 days from Colorado and return the results to Toronto. During the downtime I'll take the client into Network activity suspend and visit occasionally to see if the front page message changes to go back online. Ditto .  |
||
|
|
|