Index  | Recent Threads  | Unanswered Threads  | Who's Active  | Guidelines  | Search
 
World Community Grid Forums
  Category: Support
  Forum: BOINC Agent Support
  Thread: [Completed] Extended BOINC Server Outage Planned from 21:30 UTC Jan 13 thru 8:00 UTC Jan 14th

Quick Go »
No member browsing this thread
Thread Status: Active
Total posts in this thread: 76
Posts: 76   Pages: 8   [ Previous Page | 1 2 3 4 5 6 7 8 | Next Page ]
[ Jump to Last Post ]
Post new Thread
Author
Previous Thread This topic has been viewed 96712 times and has 75 replies Next Thread
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Extended BOINC Server Outage Planned from 21:30 UTC Jan 13 through 8:00 UTC Jan 14th
You're right, but I was expecting everything to upload, not just the scientific results, but after the brief deferral, the workunit indeed went.
[Jan 14, 2012 9:24:22 AM]   Link   Report threatening or abusive post: please login first  Go to top 
JmBoullier
Former Community Advisor
Normandy - France
Joined: Jan 26, 2007
Post Count: 3716
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: Extended BOINC Server Outage Planned from 21:30 UTC Jan 13 through 8:00 UTC Jan 14th
OK.
In this Distributed Computing activity "uploading" is used only for those scientific result files.
Reporting completed WUs, requesting new ones etc... is done via a scheduler request which is practically the client sending its client_state control file and receiving it back updated by the scheduler server. Only one exchange of a single file, and with a different server than the data server.

If that may help make things more clear. smile
----------------------------------------
Team--> Decrypthon -->Statistics/Join -->Thread
[Jan 14, 2012 10:14:26 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Extended BOINC Server Outage Planned from 21:30 UTC Jan 13 through 8:00 UTC Jan 14th
Have closed state of this thread with confirming that most work with quorum has been passed through the validators.

Till next time... crunching on.

--//--
[Jan 14, 2012 10:34:59 AM]   Link   Report threatening or abusive post: please login first  Go to top 
JmBoullier
Former Community Advisor
Normandy - France
Joined: Jan 26, 2007
Post Count: 3716
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: Extended BOINC Server Outage Planned from 21:30 UTC Jan 13 through 8:00 UTC Jan 14th
confirming that most work with quorum has been passed through the validators.
Euh, not really... but there is still one hour to go before the midday stats. wink

Edit: Everything validated now.
----------------------------------------
Team--> Decrypthon -->Statistics/Join -->Thread
----------------------------------------
[Edit 1 times, last edit by JmBoullier at Jan 14, 2012 11:51:41 AM]
[Jan 14, 2012 11:03:18 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Extended BOINC Server Outage Planned from 21:30 UTC Jan 13 through 8:00 UTC Jan 14th
every one of mine has been returned but every one of em are pending validation....

all vallidated now
----------------------------------------
[Edit 1 times, last edit by Former Member at Jan 14, 2012 11:32:39 AM]
[Jan 14, 2012 11:06:06 AM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Extended BOINC Server Outage Planned from 21:30 UTC Jan 13 through 8:00 UTC Jan 14th
The planned outage: signed, sealed, and delivered by WCG. Well planned, well executed, well delivered, and finally -- well received. Next stop: GPU Beta! wink
;
[Jan 14, 2012 12:27:03 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Extended BOINC Server Outage Planned from 21:30 UTC Jan 13 through 8:00 UTC Jan 14th
On performance improvement with MySQL 5.5, looked in late last night and noticed that the My Grid > Stats History part was already update at 00:14 AM and the My Grid > Project detail section at 00:29 AM (before the lock out). Just now saw that the My Grid > Stats History was updated with noon data at 12:11 and Project Detail at 12:25... for the new stats hungry cool

--//--

edit: The catchup from the outage did not show up on Saturday, rather just now with an estimated 50 years extra for a Sunday morning in January 2012.

228:267:05:26:55 252,423,545 471,705
----------------------------------------
[Edit 1 times, last edit by Former Member at Jan 15, 2012 12:52:43 PM]
[Jan 15, 2012 12:36:44 PM]   Link   Report threatening or abusive post: please login first  Go to top 
knreed
Former World Community Grid Tech
Joined: Nov 8, 2004
Post Count: 4504
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: Extended BOINC Server Outage Planned from 21:30 UTC Jan 13 through 8:00 UTC Jan 14th
The performance of the database has been very nice. In fact it has exceeded my expectations. There are a number of things that were running so much faster that I've been going through log files to make sure that they are actually running correctly. This is definitely going to allow us to handle considerably more load.
[Jan 15, 2012 1:36:49 PM]   Link   Report threatening or abusive post: please login first  Go to top 
knreed
Former World Community Grid Tech
Joined: Nov 8, 2004
Post Count: 4504
Status: Offline
Project Badges:
Reply to this Post  Reply with Quote 
Re: Extended BOINC Server Outage Planned from 21:30 UTC Jan 13 through 8:00 UTC Jan 14th
We will see the effect from the outage over the next few days. If you look at the chart below, this is the # of results returned by hour over the last 24 hours (1 means it was returned 1 hour ago). It looks like most computers were deferred for 24 hours (mine were - I forgot to increase the buffer on mine!) and you can see when they started to communicate with our servers again:


+----------+--------------------------------------------------+
| count(*) | round((unix_timestamp() - received_time)/3600,0) |
+----------+--------------------------------------------------+
|    17179 |                                                0 |
|    34514 |                                                1 |
|    33914 |                                                2 |
|    33003 |                                                3 |
|    33133 |                                                4 |
|    37117 |                                                5 |
|    20392 |                                                6 |
|    13793 |                                                7 |
|    22124 |                                                8 |
|    20046 |                                                9 |
|     7879 |                                               10 |
|     9102 |                                               11 |
|    17124 |                                               12 |
|    11414 |                                               13 |
|     8463 |                                               14 |
|     7824 |                                               15 |
|     8077 |                                               16 |
|     8322 |                                               17 |
|     8087 |                                               18 |
|     8098 |                                               19 |
|     7738 |                                               20 |
|     7945 |                                               21 |
|     8318 |                                               22 |
|     7429 |                                               23 |
|     3511 |                                               24 |
+----------+--------------------------------------------------+
[Jan 15, 2012 1:43:08 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Former Member
Cruncher
Joined: May 22, 2018
Post Count: 0
Status: Offline
Reply to this Post  Reply with Quote 
Re: Extended BOINC Server Outage Planned from 21:30 UTC Jan 13 through 8:00 UTC Jan 14th
Wish the client had an internal override function to basically say: I've got an idle core... please go try anyhow.

Various mechanisms have been discussed in this thread [an "how to" included on setting the client to autonomously resume after the expected outage] and some suggestions such as the WCG server bouncing any client that connects hours prior to the outage. Just thinking, a redirect mechanism that steps in to bounce clients for a preset amount of time.

Both the Known Issues and this sister thread are demonstrating the low reach. For such rare longer scheduled outs, I'd personally not hesitate to have pushed it into the RSS feed that connects into the 6.12 and higher clients, even a mail to those that have had results returned in the past 7 days. Whilst, the effective loss of contributed time seems to have stayed well in bounds. Think by tonight's stats we can more firmly assess what it worked out at.

Of course, those had a backup project defined [I have 2] would have kicked in would things have lasted way beyond scheduled times.

Of course with that "better than you hoped for" performance hike, pretty please, whilst the capacity is not yet consumed by growth, size down tasks for some of the sciences such as FAAH and HPF2 that now sit in the 10 hour mean bracket. Their UL/DL bandwidth is small, i.e. would not cost WCG an arm and a leg to do. Whilst I write this, would not be surprised if you had that already in the planner.

cheers and many thanks for achieving this step so smoothly [from our outside perspective]

--//--
----------------------------------------
[Edit 1 times, last edit by Former Member at Jan 15, 2012 3:25:35 PM]
[Jan 15, 2012 3:24:21 PM]   Link   Report threatening or abusive post: please login first  Go to top 
Posts: 76   Pages: 8   [ Previous Page | 1 2 3 4 5 6 7 8 | Next Page ]
[ Jump to Last Post ]
Post new Thread